volume XXVII, 1835, page 166, is an account, by Knight, of the

discovery, at Jamestown, of what were probably two teeth of a bison in a fragment of the jaw. These were encountered by John Hazeltine, in digging for a foundation of a building at the outlet of Chautauqua Lake, and at a depth of 10 feet. The soil was mostly gravel, but the jaw was said to have been lying in black muck. It was sent to Yale College, but was not recognized as belonging to _Bison_. Reasons were suggested why it did not belong to a young mastodon. The measurements given of the teeth agree well with the upper molars of an American buffalo. Joseph Leidy (Jour. Acad. Nat. Sci., Phila., vol. VII, 1869, p. 371) quoted Knight’s account as indicating a buffalo. The discovery is interesting, taken in connection with the finding of the specimen at Syracuse.

NEW JERSEY.

(Map 27.)

1. _Trenton, Mercer County._—Mr. Ernest Volk (Papers Peabody Mus., vol. V, 1911, p. 209, plate CXX) reported the discovery of a part of a femur of _Bison_ (probably _B. bison_) in the “yellow drift,” at Trenton, 2.5 feet from the surface. A first right upper molar, identified as that of _Bison_, was found in another sand-pit at a depth of 9 feet (op. cit., p. 136). This appears to have belonged in the Trenton gravel, but at that point the materials were apparently a mixture of sand and loam. The reader is referred to page 304, where the geology of this locality is described and a list of the species is given.

PENNSYLVANIA.

(Map 27.)

1. _Stroudsburg, Monroe County._—In Crystal Hill (Hartman’s) Cave, near Stroudsburg, was found a lower jaw containing the last molar, as noted by Leidy (Proc. Acad. Nat. Sci., Phila., 1880, p. 347; Ann. Rep. Geol. Surv. Pennsylvania for 1887, p. 5). Mercer (Proc. Acad. Nat. Sci. Phila., 1894, p. 98), mentions a tooth of the existing bison found in Hartman’s Cave.

2. _Riegelsville, Bucks County._—From a cave near Riegelsville, was sent to the Academy of Natural Sciences of Philadelphia, more than 70 years ago, a collection of bones, reported on by Leidy in 1880 (Proc. Acad. Nat. Sci., Phila., 1880, p. 349) and in 1889 (Ann. Rep. Geol. Surv., Pennsylvania, 1887, pp. 18–19). In the contribution of 1880, Leidy included _Bison_ among the animals represented, but this is not included in the list of 1887. Why this was omitted is not known. If _Bison_ occurred there, the probability is that it was represented by the existing buffalo.

INDIANA.

(Map 27.)

1. _Jasper County._—The only record known to the writer of the finding of buffalo bones worthy to be regarded as fossil is that of the former State geologist, John Collett (Geol. Surv. Indiana, vol. XII, p. 73), who makes the statement that in Jasper County bones of the buffalo, the beaver, and the bear are common.

ILLINOIS.

(Map 27.)

1. _Sullivan, Moultrie County._—In 1875 (Geol. Surv. Illinois, vol. VI, p. 186), the geologist George C. Broadhead reported that he had found the skull of a bison on the west bank of Kaskaskia River, about 3 miles southeast of Sullivan, on the land of John Purvis. The locality appears, therefore, to have been somewhere near the south half of the eastern line of township 13 north, range 5 east. The summit of the bluff here is described as rising about 25 feet above the stream. At the height of about 8 feet was a bench approximately 10 feet wide, and the skull was found on this bench, “a few feet from the top.” The surrounding clay was described as being a rich black loam.

Broadhead stated that the skull measured 12 inches across the forehead above the eyes and the same between the roots of the horns. The latter were short, thick, and slightly curved. In the Transactions of the St. Louis Academy of Science, volume III, page XXIII, practically the same account is given of the discovery. Here Broadhead expressed the idea that the skull belonged to _Bison latifrons_, and said that the horns were short, thick, and curved upwards and forwards. It is not known where the skull now is. To the writer it appears most probable that the skull was that of _Bison bison_. There is nothing in the description to indicate any of the other known species. As to the age of the deposits, the presumption is reasonable that they belong to the Late Wisconsin or Recent, for the locality is north of the Shelbyville moraine. It is possible that the bench belongs to the Illinoian; but the nature of the material, “a rich black loam,” seems to show that the bench is an alluvial deposit laid down since Wisconsin times.

2. _Homer, Champaign County._—In the collection at the State University of Illinois, at Champaign, are the horn-cores and the rear of the skull of _Bison bison_, reported to have been thrown out of a ditch near Homer. The writer is informed by Professor R. M. Bagg, of Appleton, Wisconsin, that the specimen was found in excavating a ditch, at a depth of 4 feet, according to the report made to him. Homer is situated on a part of the Champaign moraine and the bison in question must be not older than Late Wisconsin. If it was really found at a depth of 4 feet it would seem to date well back in the Recent, if not into the Pleistocene.

3. _Niantic, Macon County._—Professor A. H. Worthen reported (Geol. Surv. Illinois, vol. V, p. 308) the presence of bones of the buffalo in an old filled-up marsh near Niantic. The situation is more particularly described on page 102. With the bison bones were found those of the mastodon, the elk, and the Virginia deer. The bones of these animals are said to have been found under 4 feet of black muck, partly embedded in a light-gray quicksand filled with shells of _Planorbis_, _Cyclas_, and _Physa_.

Inasmuch as Niantic is situated near the border of the Shelbyville moraine, all these remains probably belong to Late Wisconsin times. It would be useful to know whether the bones of the buffalo, the elk, and the deer were found above those of the mastodons or mingled with them.

4. _East of Whitewillow, Kendall County._—In township 35 north, range 8 east, probably in section 27, on land owned by John Bamford, in clearing out a well in a bog, have been found the bones of mastodons and other species of vertebrates. For a description of the locality and the species found there see page 337. Mr. George Langford, of Joliet, has reported the occurrence of bones of the existing bison there and has sent to the writer a maxilla which contained finely preserved teeth.

Unfortunately, no thorough and systematic examination of the place has yet been made. All of the species and the deposit belong to the Late Wisconsin, that part of it following the withdrawal of the ice. Mr. George Langford informed the author that he found the bison and deer bones mixed up more or less with the mastodon bones. At a depth of about 4 to 5 feet the owner of the place began to strike bones of the bison, which appeared very fresh, retaining considerable animal matter. From about 6 feet down to gravel, about 13 feet, mastodon and other bones were literally packed together.

5. _Batavia, Kane County._—Dr. E. S. Riggs, of the Department of Palæontology, Field Museum of Natural History, wrote to the author that he had picked up some bison bones along a ditch in which mastodon bones had been found; but the depth at which they had been met with could not be determined. At the same time bones of the elk were found. Undoubtedly the mastodon remains belong to Late Wisconsin times; and it is probable that the bison and elk remains are to be referred to the same.

6. _Galena, Jo Daviess County._—In the collection of the Academy of Natural Science of Philadelphia is a lower hindermost molar collected in a lead crevice somewhere near Galena. It was presented to the Academy by Mr. Henry Green, of Elizabeth, a town near Galena. This, with a metacarpal bone of _Megalonyx jeffersonii_, had been found at a depth of 130 feet from the surface. It was described and figured by Leidy (Contributions to Extinct Vert. Fauna, etc., 1873, p. 255, plate XXXVII, fig. 4). Leidy thought that it might have belonged to _Bison bison_, but not improbably to _B. latifrons_. J. A. Allen (The American Bisons, etc., p. 13) concluded that it belonged undoubtedly to the existing American species. The structure of the tooth will apparently not decide this matter. It is probable that most of the animals found in those lead crevices belong to pre-Wisconsin times; and the tooth in question may belong to an extinct species. A list of the species found in the lead region of Illinois, Iowa, and Wisconsin is to be found on page 343.

7. _Mitchell, Madison County._—In “Records of Ancient Races in the Mississippi Valley” (1887), William McAdams, of Alton, Illinois, stated that in a large mound, square in shape, 300 feet on each side and 30 feet high, through which the railroads pass in the American bottom, at Mitchell, had been found, in contact with a number of copper implements and ornaments, a number of teeth of the buffalo. These McAdams had in his possession. While these teeth can not be regarded at all as belonging to Pleistocene times, the fact is of interest in connection with McAdams’s statement that in all his explorations during a period of more than 30 years, in no other case had he been able to find any evidences of the buffalo associated with the remains of the ancient people of this country. In this connection may be considered Shaler’s views on the modern coming of the buffalo east of the Mississippi River. On the other hand, account must be taken of the finding of a skull of a buffalo deep in lake deposits at Syracuse, New York.

WISCONSIN.

(Map 27.)

1. _Bluemounds, Dane County._—In his report, made in 1862, on the geology of the lead region of Wisconsin (Geol. Surv. Wisconsin, vol. I, p. 136), J. D. Whitney recorded the finding of bison bones in a crevice at Bluemounds. From the same crevice were obtained bones and teeth of the mastodon and of a peccary, and bones of a wolf. It was supposed that these remains were found at a depth of about 40 feet and embedded in the red clay commonly found in such crevices. These bones were put into the hands of Jeffries Wyman for identification, who, on page 421, stated that the bison bones were all of the size of the same parts of the existing buffalo and closely resembled them. J. A. Allen (Amer. Jour. Sci., vol. XI, 1876, p. 47), in referring probably to the same bones, speaks of “an extinct bison,” without, however, giving any reasons for his conclusion. It is nevertheless possible that he was correct.

The writer formerly believed that the fossil vertebrates, collected in the fissures in the lead region, had lived after the close of the Wisconsin glacial stage. It seems now more probable that they belong to a pre-Wisconsin time.

2. _Oshkosh, Winnebago County._—The writer has received from Dr. S. Weidman, State geologist of Wisconsin, a humerus, found in a marsh near Oshkosh, quite evidently that of _Bison bison_. Although stained by iron on the outside, the remainder of the bone is white and full of animal matter. The animal may have lived during the Recent period.

KENTUCKY.

(Map 27.)

1. _Bigbone Lick, Boone County._—Great numbers of individuals of _Bison bison_ have been found at Bigbone Lick. Cooper (Monthly Amer. Jour. Geol., vol. I, pp. 207, 211) reported numerous bones of buffaloes and even an entire skeleton, but they appear to have been near the surface or even on it. Lyell (“Travels in North America,” Murray’s ed., vol. II, p. 65) stated that he had seen great quantities of remains of the bison in a superficial stratum in the river bank; but he was left in doubt whether or not the animals had been contemporaneous with the mastodon. Shaler (Geol. Surv. Kentucky, n. s., vol. III, p. 197) found abundant remains of the buffalo at this place; but the bones were not found at any great depth, except in the bog about the spring. He regarded it as proven that the musk-ox and the caribou did not come into contact with the recent buffalo, but were extinct before it came. Some of the bison materials collected by Shaler were described by Dr. J. A. Allen, in 1876 (Mem. Mus. Comp. Zool., vol. IV; Mem. Geol. Surv. Kentucky, vol. I, pt. 2). It may be difficult to prove that any of the bison bones and teeth found here are of Pleistocene age; but there appears to be no good reason why this species might not have reached that region at the close of the Wisconsin ice-stage. A list of the species of mammals found here is given on page 403.

2. _Bluelick Springs, Nicholas County._—In the mass of materials collected in the spring at Bluelick Springs by Mr. Thomas W. Hunter, were skulls and parts thereof, teeth, limb-bones, and vertebræ. The actual geological age of these remains can not be established; but they were of probably late Wisconsin age.

FINDS OF CASTOROIDES IN PLEISTOCENE OF EASTERN NORTH AMERICA.

NEW YORK.

(Map 28.)

1. _Clyde, Wayne County._—A skull of the giant beaver was found, about the year 1846, near Clyde, on the farm of Gen. W. H. Adams. The locality and the geological conditions were described by James Hall (Proc. Boston Soc. Nat. Hist., vol. II, 1846, p. 167; Boston Jour. Nat. Hist., vol. V, p. 385). The region is on the divide between the streams flowing north into Lake Erie and those flowing southward into Clyde River. The actual spot was at the head of a shallow stream which flows into Lake Ontario. At this point the Sodus Canal was cut and ran in a north-and-south direction. The farm was only partly swampy. Hall’s section is as follows from above downward:

1. Vegetable soil, 2 feet or more.

2. Fine sand, with some alternating layers of clay, containing twigs, leaves, etc., 2 to 3 feet.

3. Muck, or peaty soil, with decayed wood, bark, leaves, and even trunks of large trees, about 4 feet.

4. Fine sand, with fresh-water shells, 2 to 3 feet.

5. Drift, with boulders; depth unknown.

The skull was found at the bottom of No. 3, at a depth of 8 feet. It is evident that this animal lived here near, or after, the close of the Wisconsin stage, and after the old Lake Iroquois had withdrawn from the region.

2. _Canastota, Madison County._—In 1914, Dr. Burnett Smith, of Syracuse University, reported (Amer. Jour. Sci., vol. XXXVIII, p. 463) the discovery, at this place, of an incisor tooth of the giant beaver. The exact locality is given as about 225 paces northwest from the southeast line of lot 10, town of Lenox, on Cowaselon Creek, otherwise known as the “State ditch.” The tooth was found at a depth of 9 feet, in a sticky blue clay, containing a few fresh-water shells. Just above this, at a depth of 7 feet, is a layer made up principally of shells, with some vegetable matter. This animal could not have lived here until after the withdrawal of Lake Iroquois, and therefore not till near the close of the Wisconsin stage.

PENNSYLVANIA.

(Map 28.)

1. _Stroudsburg, Monroe County._—In 1889, Dr. Joseph Leidy reported (Ann. Rep. Geol. Surv. Pennsylvania, 1887, p. 14, plate II, figs. 7–20) the discovery of teeth of _Castoroides ohioensis_ in Hartman’s (or Crystal Hill) Cave, about 3 miles southwest of Stroudsburg and 5 miles from Delaware Water Gap. Its elevation is about 800 feet above the level of Delaware River. The species associated with this giant beaver will be listed on page 309. The parts figured by Leidy are a portion of a palate, with the molars and some of the premolars, and both rami of the lower jaw, showing the three temporary molars and the first true molars, with some incisors and the permanent canines.

OHIO.

(Maps 28, 29, 36.)

1. _Nashport, Muskingum County._—In 1836 (Amer. Jour. Sci., ser. 3, vol. XXXI, pp. 79–83), S. P. Hildreth, in an unsigned article, gave an account of the finding of remains of the type specimen of the giant beaver, in association with remains of mastodon and of a supposed fossil sheep, at a point 2 miles north of Nashport. A canal, now abandoned, was being constructed, which followed two small streams, one of which flowed into Licking River, the other into Wakitomika Creek. The land traversed was flat and swampy. The distance from Nashport to Wakitomika Creek is nearly 4 miles, so that in saying that the spot was on this creek Hildreth spoke in general terms. The bones of the mastodon and the right halves of the lower jaws of two giant beavers were found resting on a bed of gravel at a depth of 14 feet. Foster (2d Ann. Rep. Geol. Surv. Ohio, 1838, p. 80) stated that a molar and a tusk of an elephant had also been found here. Hildreth concluded that the jaws and teeth were perhaps those of an animal of the beaver family; “or, from the grooved outer surfaces of the incisors, a marine animal of the walrus or seal race, and a borderer of the ancient ocean.” It was afterwards described by J. W. Foster (2d Ann. Rep. Geol. Surv. Ohio, 1837, p. 80, figs.) under the name of _Castoroides ohioensis_. The remains described consisted of the front end of one side of a lower jaw with its incisor, an upper incisor, and a radius. They showed signs of some attrition; but in a region like that they could not have been transported any considerable distance.

In the mud in which the canal at this point was cut, there were found three skulls of a species of sheep, which Hildreth thought were different from those of the domestic sheep and to which he gave the name of _Ovis mamillaris_. They are said to have been discovered at a depth of 8 feet. It seems quite possible that they had been lying on or near the surface and had made their way to the side of the canal by the flow of the mud, which gave much trouble by filling up the canal during the night. Most, if not all, of the differences thought to separate these skulls from the domestic sheep disappear on comparison. The specimens of both _Castoroides_ and of the sheep have probably been lost. They appear not to be at Zanesville. On page 82 of the article above cited, Hildreth stated that he had received, from some point on Wills Creek, a portion of a tooth similar to the one found at Nashport; the place was said to be about 40 miles east, apparently, of Zanesville. This would seem to be in Noble County. The tooth was described as being embedded in dark-colored carbonate of lime and as having fallen from a calcareous rock which lies near the tops of the hills, 150 feet above the bed of the creek. It is very probable that this was not a tooth of _Castoroides_. It may have been the spine of a palæozoic shark.

2. _Wilmington, Clinton County._—From Professor W. C. Mills, of the Ohio State University, the writer in 1913 obtained information that a fine skull of _Castoroides_, without the lower jaw, had been found on the farm of Mr. J. M. Richardson, on the western border of Wilmington. Nothing more has been learned about the discovery. The locality is north of the Hartwell moraine, and the animal must have lived there after the withdrawal of the ice-sheet from that region.

3. _Germantown, Montgomery County._—One mile east of Germantown, Edward Orton, State geologist of Ohio, found along Twin Creek a large tooth which (Amer. Jour. Sci., ser. 2, vol. L, 1870, p. 54) he compared with the tusk of a hog. It was later identified by J. S. Newberry (Proc. Lyc. Nat. Hist. New York, vol. I, 1870, p. 83) as belonging to _Castoroides_. It was found in a bed of peat which is overlain by from 50 to 100 feet of glacial drift. One might conclude that the animal had lived there at some time between the Illinoian and Wisconsin stages. However, opinions have differed.

The geology along Twin Creek has been studied by Orton, Wright, and Leverett. The last named published his views in 1902 (Monogr. U. S. Geol. Surv., XLI, pp. 363–365, plate XIV, fig. 1). He states (p. 365) that there seem to be good reasons for believing that the peat-bed indicates the lapse of a considerable interval of deglaciation. Whether the interval preceded or followed the formation of the early Wisconsin moraine is yet to be determined. That seems to mean that the interval may be mid-Wisconsin or pre-Wisconsin. Wright thought that but a few hundred years had elapsed between the deposit of the till below the peat and that above. Orton’s description of the locality was published in 1870 (Amer. Jour. Sci., ser. 2, vol. L, p. 54).

4. _West Sonora, Preble County._—In 1893 (Amer. Geologist, vol. XII, p. 73), Professor Joseph Moore reported that a fragment of an upper incisor of _Castoroides_ had been found at West Sonora. It was associated with remains of a mastodon. West Sonora is on the Englewood moraine.

5. _Greenville, Darke County._—In 1883 (Jour. Cin. Soc. Nat. Hist., vol. VI, p. 238), F. W. Langdon described a tooth of _Castoroides_, found at a depth of 4 feet, in a swampy locality near Greenville. In 1893 (Amer. Geol., vol. XII, p. 73), Joseph Moore stated that this tooth belonged to Dr. J. W. Jay, of Richmond. It may now possibly be in the collection of Earlham College. Moore said that it had been found associated with mastodon.

In the public library at Greenville is a fragment of an upper incisor of _Castoroides_, found in making a ditch along Bridge Creek, in 1889, by Mr. Leo Katzenberger, who writes that the place is in the northwest corner of section 1, township 11, range 2 west, 1.5 miles southwest of Greenville. These animals likewise lived on or near the Sidney moraine.

6 _New Knoxville, Auglaize County._—In C. W. Williamson’s “History of Ohio and Auglaize County,” 1905, on page 338, with a figure, is an account of the finding of a skull of _Castoroides ohioensis_ in section 29 of Washington Township, which is in township 6 south, range 5 east, and near New Knoxville. The discovery had been made that beneath a bed of humus there was a stratum of gravel of a quality for road making. In removing the upper peaty layer, the head of the giant beaver was discovered, near the south margin of the pond. Williamson stated that the house of the animal was uncovered. It was between 3 and 4 feet high and about 8 feet square; the poles of which it was constructed were about 3 indies in diameter and were laid after the manner of the houses of modern beavers. Apparently the beaver died in the house, and it was thought that after the death of the beaver wolves or other carnivorous animals had inhabited the house, since bones of deer and other animals were strewn over the floor. It is to be regretted that the house, if such it was, was not taken up in a way that it might have been accurately reconstructed. Williamson’s account is reproduced in Bulletin 16, Geological Survey of Ohio, 4th series, 1912, page 39.

In Heidelberg University, Tiffin, Ohio, the writer has seen a very large skull of _Castoroides_, labeled as found at Wapakoneta, but it is quite certainly the one found at New Knoxville. Both incisors are broken off close to their insertion in the skull. Williamson’s figure represents at least the left one present.

MICHIGAN.

(Map 28.)

1. _Berrien County._—In the American Museum of Natural History, New York, is a nearly complete skull with the left ramus of the lower jaw, purchased from Mr. George A. Baker. The exact place in the county where it was found is unknown, and the writer has been unable to get into communication with Mr. Baker.

As to the time in the Pleistocene when this individual lived, we may be sure that it was after the Wisconsin glacial ice-sheet had abandoned this county. How long after this retirement it is impossible to say. It is to be noted that both mastodons and mammoths have been found in this county, in what appear to be deposits of the same age.

2. _Adrian, Lenawee County._—In the U. S. National Museum is a skull of _Castoroides_ (Cat. No. 197), of which the lower jaw is missing. This was received June 10, 1880, from Professor J. Kost, then of Adrian College, Michigan. In his letter Professor Kost wrote as follows:

“Found in fresh-water marsh, 4 feet under, in Adrian, Lenawee Co., Michigan. In same place as the Decker mastodon, now in Adrian College; also of lower jaw of smaller mastodon (sent in this consignment), with various bones of elk, deer, etc.”

The mastodon jaw referred to is in the U. S. National Museum (No. 188). The present writer has not been able to learn exactly where all these bones were obtained. It would be interesting to know whether all–mastodons, giant beaver, elk, and deer–were found in the same excavation. It is probable that they were at least in nearly the same spot. For remark on the age of the deposits at Adrian see page 81.

3. _Ann Arbor, Washtenaw County._—In the collection of the Department of Geology in the University of Michigan is a skull which lacks the lower jaw and is otherwise slightly injured. A report of this specimen was made in 1914 by Mr. N. A. Wood (Science, n. s., vol. XXXIX, p. 759). This was found several years ago in a peat-bog on the farm of Professor J. B. Steere, 3 miles south of Ann Arbor, at a depth said to have been about 3 feet. Beneath the peat and muck is a gravelly marl. According to the Ann Arbor Folio (No. 155, U. S. Geol. Surv.), there is, running south from the city, a strip of low ground designated as occupied by peat and muck. This borders on the east a part of the Fort Wayne moraine, and must have provided an ideal spot for colonies of these great beavers. Naturally these specimens must be credited to the Late Wisconsin stage.

4. _Attica, Lapeer County._—In the collection of Alma College, Alma, Michigan, is a fragment of an upper incisor, found at a depth of 7 feet, in digging the tail-race of a mill in Attica. The statement was made that at the same place there were often found what appeared to have been beaver dams made of wood. This wood crumbled on coming to the air. In cases like this there is a fine opportunity to determine whether or not the wood had been gnawed by the broad incisors of _Castoroides_ or by the narrower ones of the existing beaver. The wood might easily be prevented from crumbling by replacing the water with a solution of gum arabic or even of glue.

Attica is situated some distance outside of the beaches of old Lake Maumee, and on low ground between morainic tracts left by the Saginaw lobe in its retreat. These gigantic beavers must, therefore, have lived near the close of the Pleistocene.

5. _Owosso, Shiawassee County._—In the collection of the University of Michigan (No. 3109) is the greater part of a lower jaw of a giant beaver, found somewhere near Owosso, in a swamp deposit. An account of this specimen was given in 1914 by Mr. N. A. Wood (Science, n. s., vol. XXXIX, p. 758). It was received from Mr. A. G. Williams in 1892. According to Leverett and Taylor’s glacial map of Michigan, Owosso lies a few miles outside of the beach of old Lake Saginaw. This is supposed to have come into existence about the close of the period of Lake Maumee. The earliest time when this beaver might have existed, leaving out the question of the climate, would coincide closely with the time when the one found at Attica might have lived. It is most probable that both lived at a time when the glacier front was farther away.

INDIANA.

(Maps 28, 30.)

1. _Vanderburg County._—In 1884 (14th Ann. Rep. Geol. Surv. Indiana, pt. 2, p. 37), in a footnote written probably by John Collett, State geologist, it is stated that remains of _Castoroides ohioensis_ had been found in this county. Inasmuch as this county lies outside of the drift region, and as no details as to place and depth were given, we can arrive at no conclusion as to the stage of the Pleistocene in which the possessor of this tooth lived. The reader may consult page 258.

2. _Richmond, Wayne County._—About 2 miles east of Richmond, where a farmer was scooping out wet earth for a fish-pond, there was found by Joseph Moore (Amer. Geologist, vol. XII, p. 73) a fragment of an upper incisor of this species. With it were sound and decayed teeth of the mastodon. Most probably this fish-pond was being excavated in low ground where a marsh had existed. Richmond is situated just south of the Bloomington moraine, on an area which is undulating and more or less morainic. The animal must have lived at some time after the culmination of the Wisconsin stage.

3. _Greenfield, Hancock County._—In 1893 (Amer. Geologist, vol. XII, p. 73), Joseph Moore mentioned the fact that some remains of _Castoroides_ had been found near Greenfield and that these were in the possession of Dr. M. M. Adams. In 1900 (Proc. Ind. Acad. Sci. for 1899, p. 171, plates I, II), Moore presented figures of the skull and made some brief statements regarding it. At that time the skull had come into the possession of Earlham College. If restored this skull would have had a length of 13 inches. Nothing is known as to the exact place where it was found, but it can not be doubted that the animal lived after the Wisconsin ice had retreated further north.

4. _Jamestown, Boone County._—In the State Museum at Indianapolis is a lower jaw of a giant beaver which has all of the molars, but whose incisors are broken off at the border of the bone. This specimen was presented by Mr. A. E. Deatley, of Lizton, Hendricks County, who found it in earth thrown out by a dredging machine, but the exact locality was not stated. Jamestown is situated on Eel River where it crosses the Champaign moraine. The geological age of the animal is therefore Late Wisconsin.

5. _Summitville, Madison County._—In the State Museum at Indianapolis is an upper right incisor of the giant beaver in its premaxilla, labeled as presented by Mr. J. F. Cartwright. Nothing more is known of the history of the specimen.

Summitville is surrounded by plains of Wisconsin drift. It is about 12 miles from the place where was found the fine mounted specimen of _Elephas primigenius_ now in the American Museum of Natural History, New York.

6. _Union City, Randolph County._—Here was found the nearly complete skeleton of _Castoroides ohioensis_ at Earlham College, Richmond, Indiana. This was secured by Professor Joseph Moore, who described and figured it. It was discovered on the farm of John M. Turner, about 8 miles nearly east of Winchester. Mr. Turner has informed the writer that the farm is a part of section 15, township 17, range 1.

The bones occurred in a layer of fine-grained marly silt from 2 to 3 feet thick, overlain by from 3 to 4 feet of dark loose mold abounding in fragments of shrubby stems and vines in various stages of decay. Under the silt containing the bones were coarser and finer drift gravels which formed the bottom of the ditch. In the silts were found fresh-water gasteropods and bivalve shells. Along the same ditch, within a distance of 30 rods, other fragments were found which were supposed to indicate 9 individuals of _Castoroides_. As this region is covered by Wisconsin drift, the animal evidently lived after the Wisconsin ice-sheet had retired from the Union City moraine, possibly a long time thereafter.

7. _Fairmount, Grant County._—Near Fairmount were found some limb-bones and other parts (but no skull) of the giant beaver. These were obtained not far from where the large specimen of _Elephas primigenius_ was discovered which is mounted in the American Museum of Natural History in New York. The remains of this _Castoroides_ are in the Field Museum of Natural History. No details regarding the find have been published. It was stated that near the bones were parts of trees, as though a dam had been built there; but this interesting matter appears not to have been investigated.

The elephant mentioned above was found on the farm of Dora C. Hitt, in the southeast quarter of section 23, township 23 north, range 8 east.

8. _Carroll County._—In 1884 (14th Ann. Rep. Geol. Surv. Indiana, pt. 2, p. 37) the State geologist, John Collett, wrote that _Castoroides_ had been found in this county; but nothing was added to this statement. On the map the number is placed arbitrarily.

9. _Logansport, Cass County._—In the U. S. National Museum is a fine skull of _Castoroides_, without lower jaw, which, according to the newspaper report accompanying it (dated January 30, 1894), was found 2 or 3 miles north of Logansport, by Mr. S. L. McFadin, who sold it to the National Museum. It lay at a depth of 7 feet on a fine sand, above which was a foot of solid gravel, then 3 feet of solid clay, and at the top 3 feet of alluvium. According to Leverett and Taylor’s map of the region (Monogr. U. S. Geol. Surv., vol. LIII, plate VI), this place would be on the moraine which lies north of the Wabash River, the meeting-place of the ice-lobes coming from Lake Michigan, Lake Erie, and Saginaw Bay.

10. _Macy, Miami County._—From Mr. C. F. Fite, Denver, Indiana, the writer received a photograph of a tooth of _Castoroides_, apparently the lower right incisor. This was found in Allen Township. Mr. Fite gives as the exact locality section 23, township 29, range 3 east. This would be not far from Macy. It lies, therefore, on or near the northern border of the great moraine which extends from Delphi, Indiana, to the northeastern corner of the State.

11. _Kosciusko County._—As in the case of Cass County, we depend for our knowledge of the discovery of _Castoroides_ in Kosciusko County on the statement made by John Collett, in the place there cited.

12. _Grovertown, Starke County._—From Dr. E. S. Riggs, of Field Museum of Natural History, the information has been received that there is at that museum a fine skull, with the right half pf the mandible, of a giant beaver which was found 1.5 miles west of Grovertown, in making an excavation for the abutment of a bridge, 6 feet below the surface in township 34 north, range 1 west. This is within the region of the Pleistocene Lake Kankakee.

ILLINOIS.

(Maps 28, 38.)

1. _Shawneetown, Gallatin County._—In the collection of the Academy of Natural Science of Philadelphia are a part of one incisor, two molars, and two petrous bones which were many years ago obtained by a Dr. Feuchtwanger, from a well at a depth of 40 feet. These were mentioned by Le Conte in 1852 (Proc. Acad. Phila., vol. VI, p. 53). Leidy has figured the incisor (Holmes’s “Post-Pliocene Fossils of South Carolina,” 1860, plate XXII, fig. 5; Ann. Rep. Geol. Surv. Pennsylvania, 1887, plate II, fig. 10). Leverett (Monogr. U. S. Geol. Surv., vol. XXXVIII, p. 65) states that at Shawneetown a boring for gas and oil penetrated 112 feet of alluvial and other deposits before reaching rock. His map (plate VI) indicates that here the valley of the Ohio is composed of sand and gravel plains of Wisconsin age. Under the conditions it seems impossible to form any certain conclusions regarding the geological age of this specimen. It belongs possibly to the later half of the Pleistocene.

2. _Alton, Madison County._—In the McAdams collection, described on page 338, is a part of a large upper incisor, in two pieces, of a specimen of _Castoroides_, with McAdams’s Nos. 209, 210, and a small fragment of another incisor. All three specimens are more or less enveloped in nodules of hard materials. In 1883 (Trans. St. Louis Acad. Sci., vol. IV, p. LXXX) McAdams stated that he had seen, both in true and modified drift, remains of rodents large and small, but one, an extinct beaver, was of monstrous size.

For conclusions as to the age of the fauna secured by McAdams see page 339.

3. _Charleston, Coles County._—In 1867 (Proc. Acad. Nat. Sci. Phila., p. 97), Leidy briefly described a skull of _Castoroides_, sent to him for examination by Professor A. H. Worthen. It lacked both zygomatic arches and the incisor teeth. The length of the skull was 10.5 inches. This skull had been found by someone while he was plowing in a field near Charleston. The region about Charleston is covered by the Shelbyville lobe of the early Wisconsin drift. The animal must have lived at some time after the deposition of that drift.

4. _Naperville, DuPage County._—H. M. Bannister (Geol. Surv. Illinois, vol. IV, p. 113) reported a skull and other parts of the skeleton of _Castoroides_, found by a farmer in a slough not far from Naperville. The skull went to Colonel Wood’s Museum in Chicago, and it was probably destroyed in the great fire of 1871. This animal quite certainly lived after the retirement of the Wisconsin ice-sheet.

SOUTH CAROLINA.

(Map 28.)

1. _Charleston, Charleston County._—In 1860, Dr. Joseph Leidy (Holmes’s Post-Pl. Foss. South Carolina, p. 114, plate XX, figs. 6–8) recorded the fact that fragments of the teeth of the giant beaver had been found in the Pleistocene deposit of Ashley River.

In the Pinckney collection is an upper cheek-tooth, the fourth premolar. The height of the tooth is 37 mm., the length is 16 mm., the width 11.5 mm. It was found in the vicinity of Charleston.

In the Scanlan collection, the property of Yale University, and made in the vicinity of Charleston, are five more or less injured teeth. One is a left upper molar, either the second or the third. The length of the grinding-surface is 12 mm., the width 13 mm. Two fragments of upper right incisors are interesting. One of these, 140 mm. long, bears the oblique excavated surface worn by the lower incisors. Each diameter of the tooth is 25 mm. The other fragment is 123 mm. long and comes from the middle of the tooth. The two diameters of this tooth are, as in the other one, 25 mm. Both of these teeth appear to be more strongly curved than the teeth of more northern specimens. Also, the striation on the outer face of the tooth is finer, finally becoming hair-like lines as the rear face is approached. More of the larger ridges in the front of the tooth are directed obliquely and terminate along a front groove than in specimens hitherto observed. It is possible that an undescribed species is indicated. The two teeth present some differences between themselves. Another fragment, 103 mm. long, has a diameter of 20 mm. At the base is seen a part of the pulp-cavity.

GEORGIA.

(Map 28.)

1. _Brunswick, Glynn County._—In a small collection of vertebrate fossils made during dredging operations at Brunswick not many years ago, and which now belongs to the Geological Survey of Georgia, Gidley found a fragment of an incisor tooth of _Castoroides ohioensis_. The accompanying species will be recorded on page 370. Gidley’s list is found on page 436 of Bulletin No. 26 of the Geological Survey of Georgia.

MISSISSIPPI.

(Map 28.)

1. _Natchez, Adams County._—James Hall, in 1846 (Proc. Bost. Soc. Nat. Hist., vol. II, p. 168; Jour. Bost. Soc. Nat. Hist., vol. V, p. 380), announced that remains of this animal had been found in the neighborhood of Natchez. The exact locality is unknown and likewise the conditions under which the specimens were discovered. This species is not included by Leidy in his list of fossil mammals found in Pleistocene deposits in Mississippi up to 1854 (Wailles, Agri. Geol. Mississippi, p. 196).

A list of the species found in the vicinity of Natchez is presented on page 392.

TENNESSEE.

(Map 28. Figure 23.)

1. _Memphis, Shelby County._—In 1850, Dr. Jeffries Wyman reported (Proc. Bost. Soc. Nat. Hist., vol. III, p. 281) that a part of a lower jaw of _Castoroides_ had been found at Memphis. With it were a toe-bone of _Megalonyx_, a tooth of a young mastodon, and a part of the lower jaw of a beaver. It was thought that these remains had been buried in the deposits laid down by Mississippi River. It is to be regretted that the locality and the height above the river were not more exactly specified. The specimen of _Castoroides_, a right ramus of the lower jaw, is now in the Academy of Natural Sciences of Philadelphia.

ON THE PLEISTOCENE GEOLOGY OF NORTH AMERICA AND ITS RELATION TO ITS FOSSIL VERTEBRATES.

ONTARIO.

For a knowledge of the Pleistocene of Canada, the student ought first to read Dr. J. W. Dawson’s “Canadian Ice Age,” published in 1894. In this will be found references to the earlier literature of the subject. For the results of more recent studies the reports of the Canadian Geological Survey are to be consulted, as well as papers published in the scientific journals. For the more important of these papers the reader may consult the list published by Dr. H. L. Fairchild in 1918 (Bull. Geol. Soc. Amer., vol. XXIX, pp. 229).

To state the matter briefly, one may say that almost everywhere in Ontario are deposits of glacial drift of Wisconsin age. In a few localities have been discovered beds which belong to earlier glacial and interglacial epochs. On the other hand, around Hudson Bay, around the Gulf of St. Lawrence, along St. Lawrence and Ottawa Rivers, and the Bay of Fundy are marine deposits, laid down after the Wisconsin ice had retired from those localities and while the region which had been occupied by this ice-sheet was depressed so much that the sea could enter the basins named.

The most interesting locality in Canada for the student of vertebrate palæontology is doubtless Toronto, because of the presence there of Pleistocene deposits belonging to more than one stage, and because of the discovery of several species of extinct vertebrates and of many mollusks, insects, and plants. For an understanding of the geology of the region Coleman’s papers must be studied, as well as those of authors cited by him. On the interglacial deposits three of Coleman’s papers may be especially cited (Jour. Geol., vol. IX, 1901, pp. 285–310; 10th Internat. Cong. Geol., 1906, Mexico, pp. 1237–1258; Bull. Geol. Soc. Amer., vol. XXVI, 1915, pp. 243–254).

According to Coleman’s figure 1 of the first paper cited, the known interglacial deposits in that region extend from the mouth of Humber River eastward beyond the mouth of Rouge River, a distance of about 22 miles, and away from the lake a distance of about 8 miles. Deposits have been found even 14 miles north of Toronto (Coleman, 1915, p. 246). Coleman’s sketch map of the region, taken from his paper of 1901, is here reproduced (fig. 3).

According to Coleman (paper of 1915, p. 243) there are known at Toronto five well-defined sheets of boulder clay, with four sheets of interglacial sand and clay separating them. So far as the writer knows, only the lowest of these beds have been described with any particularity. These lowest beds constitute the Toronto formation, and it is these which have furnished nearly all the fossil animals and plants discovered in that region. This Toronto formation is divisible into two portions, and these have been designated as the Don beds and the Scarboro beds. They are regarded as having been deposited in the valley of an ancient river running from Georgian Bay to Scarboro. Of these the Don beds are the older. Sections of these are found in Toronto and outside, especially along Don River. They have been laid down usually on a boulder clay, 1 to 9 feet thick, which itself reposes on Hudson River shales. At one point along the Don an interglacial river had cut through both the boulder clay and the shale to a depth of 16 feet. The Don deposits consist of varying layers of sands, gravels, and clays. At one point the section obtained amounted to about 27 feet; but this, combined with another, made up about 44 feet. At one place trunks, 12 or 15 feet long, of trees have been found, which were flattened into the surface of the boulder till; also shells of unios, which are embedded in clay close to the boulder till.

In 1913 (Ontario Bur. Mines. Guide Book No. 6, pp. 15–18), Professor Coleman presented a list of the species found in the Don beds. Of the plants 32 species of trees had been secured, among them the pawpaw, the red cedar, and the osage orange; 41 species of fresh-water mollusks were listed, of which 12 were Unionidæ.

As bearing on the climate, it may be said that there are 12 species of the genus _Unio_ listed, of which 4 species are now known only from localities south of the St. Lawrence drainage; while 3 other species live in Lake Erie, but not in Lake Ontario. The plants are mostly trees; and several species, as the osage orange and the pawpaw, are now found only considerably farther south. One species of maple no longer exists. Penhallow gave it as his opinion that the flora points conclusively to the existence of climatic conditions of a character more nearly like that of the middle United States to-day. The unios now missing from that region give evidence to the same fact. For these reasons the Don deposits are spoken of as the warm-climate beds.

The Scarboro beds are finely displayed at Scarboro heights, a few miles east of Toronto. The thickness of the clay here amounts to about 94 feet. In these deposits have been found possibly mammoth or mastodon and caribou, but there is some uncertainty about these. Only 14 species of plants have been secured and these are trees; but apparently no mollusks have been reported. As an offset there are great numbers of beetles. Of these there have been described 72 species, and all are extinct except 2.

The trees, according to Penhallow, indicate a climate somewhat cooler than that now prevailing in that region. The same conclusion was reached by Scudder from his study of the insects. In his paper of 1901, Coleman took the view that the Toronto formation had been laid down in the interval between the Iowan and the Wisconsin glacial stages, that is, during what is now known as the Peorian. In the address of 1906, page 44, he appears to have been inclined to accept Leverett’s view that at least the Don beds belonged to the Sangamon stage. By 1915 (paper cited, p. 252) he had about concluded that the Toronto beds were as old as the Aftonian stage.

Dr. G. F. Wright, in 1912 (Bull. Geol. Soc. Amer., vol. XXV, pp. 205–218), accounted for the deposits and fossil animals and plants found at Toronto in a different way. At a certain time in the Pleistocene the region about Toronto was occupied by some species of animals and plants now found only considerably further south. An ice-sheet from the Keewatin center extended thither and laid down the Don beds. Later the Labrador glacier pushed into that region and deposited the Scarboro beds. According to this view the whole succession of events would be much shortened.

The writer is disposed to accept Leverett’s estimate of the geological position of the interglacial beds at Toronto. The presence there of _Elephas primigenius_, _Mammut americanum_, and the probable _Ursus americanus_ hardly counts in the determination of the geological age, for all these animals appear to have continued on from at least the Aftonian interglacial to the close of the Wisconsin. There are no specimens that show that either _Rangifer_ or _Cervalces_ existed during the Aftonian, although one can hardly doubt that they did then exist. In order to show that the Toronto formation belongs to the Aftonian, it would be necessary to produce satisfactory stratigraphical evidence or to find there genera and species of mammals which characterize the Aftonian, such as camels, _Elephas imperator_, and those horses which belong to the early Pleistocene. If the deposits belong to the Sangamon stage, such horses as _Equus complicatus_ and _E. leidyi_ ought in time to be discovered there.

Coleman has discussed the interglacial beds that occur elsewhere in Canada (10th Internat. Geol. Congr. 1906, Mexico, pp. 1237–1258; Bull. Geol. Soc. Amer., vol. XXVI, 1915, pp. 243–254). He refers to Chalmers’s account of interglacial deposits along Lake Erie; but so far as the writer has been able to determine, most of the deposits referred to are of Late Wisconsin age. However, as he says, Spencer found interglacial materials near Niagara Falls. Other beds have been discovered along Moose River, south of James Bay; but their geological position has not been definitely determined, and the fossils discovered there, mostly proboscideans, are not referred with certainty to the interglacial deposits.

Most of the vertebrate fossils found in Ontario, excepting many of those found at Toronto, belong to the Late Wisconsin stage; and in studying their geological relations one must, as in the States of New York, Ohio, Indiana, and Michigan, take into consideration the history of the Great Lakes after the Wisconsin ice-sheet began to retire. According to Leverett and Taylor’s maps (Monogr. LIII, U. S. Geol. Surv., plate XIV), as early as the time when the glacial ice had just begun to withdraw from Lakes Michigan and Erie, a considerable area of land had become cleared of ice in the peninsula bounded by Georgian Bay, Lakes Huron, Erie, and Ontario. We can hardly suppose, however, that any mastodons or any elephants, except possibly _Elephas primigenius_, could have made their way to that area. Even the last-mentioned species would have had to travel over many miles of glacial ice. Conditions were hardly more favorable when Lake Whittlesey had come into existence (op. cit., plate XVI). At a later stage (op. cit., plate XVII) the ice-free parts of the peninsula could have been reached only by crossing the lakes or over wide stretches of glacier. It is possible that some of the mastodons and elephants that have been found had crossed over into Ontario at about the stage represented by plate XIX of the work cited, but it is more probable that they lived there at a later time.

Brief mention is here made of the fossil vertebrates found in Ontario and their localities. More detailed statements will be found on the pages cited.

Beginning in the west, a mastodon has been found at Blythewood, Essex County (p. 45). In Elgin County a mastodon has been met with at St. Thomas (p. 45), and a mastodon (p. 45) and an undetermined species of elephant at Highgate (p. 45). A little farther back from the lake, at London, Middlesex County, has been found a mastodon (p. 45). At Marburg, not far from the shore of Lake Erie, Dr. H. M. Ami exhumed a mastodon (p. 45). The writer has not learned how this locality is related to the ancient beaches. At Dunnville, Haldimand County, a mastodon has been secured (p. 46). It could hardly have lived there before the lake had assumed nearly its present level. The same remark will apply to the time when the mastodon (p. 46), _Elephas columbi_ (p. 147), and possibly _E. primigenius_ (p. 166) lived at St. Catharines. From Hamilton, at the extreme western end of Lake Ontario, have been described remains of _Elephas columbi_ (p. 147), _E._ sp. indet. (p. 166), elk, _Cervus canadensis_ (p. 235), and the beaver. _Elephas primigenius_ has been found at Toronto, (p. 130); also _Cervalces_, a bison (p. 256), and a reindeer (p. 244). The same elephant has been discovered at Amaranth, in Dufferin County (p. 130). The elk, _Cervus canadensis_, has been reported from Strathroy, Middlesex County, and Kingston, Frontenac County (p. 235). At Smith’s Falls, Lanark County, the humpback whale, _Megaptera boöps_, has been discovered (p. 17). White whales, _Delphinapterus leucas_ and _D. vermontanus_, have been found at Pakenham, Lanark County (p. 17), at Cornwall, Stormont County (p. 18), Nepean Township (p. 17), Ottawa East, Carleton County, and Williamston, Glengarry County (p. 18). At Ottawa has been discovered an assemblage of species, as listed on page 287.

The geology of the Hamilton locality has been described by Logan (Geol. Canada, 1863, p. 914), by Spencer (Canad. Naturalist, vol. X, 1883, pp. 222–230, 306–308), and by Coleman (Bull. Geol. Soc. Amer., vol. XV, 1904, p. 351). The remains mentioned were found in deposits forming what is called Burlington Heights. Here Dundas Valley opens into the extreme western end of Lake Ontario. The valley is about a half mile wide. Across this had been formed a bar, interrupted only at its northern end, with a height of 108 feet above the level of the lake and a width varying from a few hundred yards to less than a half mile. Its height is almost that of the Iroquois beach found on the south shore of the lake and continuing on the northern shore. Many years ago a canal was cut through the narrowest part of the bar, and it was in the construction of this that the elephant (p. 166), elk (p. 235), and beaver bones were found. It is evident that the bones were deposited there while the bar was being built and at a time when it lacked 38 feet of being as high as it now is. The elephant jaw is in good condition, and this indicates that the animal died near the spot.

Coleman (op. cit., p. 352) stated that afterwards a railroad cut had been made across the southern end of the bar, exposing 30 feet of coarse stratified gravel, followed below by 2 feet of brown clay (evidently an old soil) and 8 feet of blue till. In the old soil were found quantities of decayed wood, as well as bones of mammoth and other animals. About a mile farther west, pits were opened for clay, sand, and gravel. Coleman gives the following geological section at this place. The column at the right gives the heights above the lake level.

_feet._ _feet._ Clay making red brick 6 78 Gravel 30 72 White sand 5 42 Hard pan 4 37 White sand with mammoth tusks and bones 33 Covered to level of the bay 0

The mammoth tusks and bones were not water-worn. It will be observed that they were found 83 feet below the top of the Iroquois beach (116 feet above the present lake), while the jaw was only about 45 feet below the beach. Both Coleman, as cited, and Fairchild (Bull. Geol. Soc. Amer., vol. XXVII, p. 247) regard the formation of the bar at Hamilton as showing that during Iroquois times the lake became flooded to a height of about 82 feet.

Besides the interglacial species found at Toronto, which have already been mentioned, there may be noted a tooth of _Elephas primigenius_ (p. 130), a cast of which was reported by Winchell. Whether this was derived from interglacial or late Wisconsin beds is not known. Coleman, as elsewhere cited, reported the finding of remains of one of the elephants on the Iroquois beach. On the same beach have been collected antlers of reindeer (p. 244). These animals must have lived there not earlier than the time when that beach was forming, perhaps later.

In a buried gorge extending in a northwestern direction from the whirlpool at Niagara to the Niagara escarpment, Dr. J. W. Spencer (Bull. Geol. Amer., vol. XXI, p. 433) has discovered what he regards as deposits equivalent to the Toronto formation, while older glacial and interglacial beds are found below and more recent ones above. No fossils were met with except wood. At Amaranth have been secured considerable parts of a skeleton of _Elephas primigenius_ (p. 130). This elephant must have existed rather late in the Wisconsin stage. About Kingston in Frontenac County, at two places, have been secured remains of the elk (p. 235), but lack of details as to places and conditions precludes certainty as to their geological age. The fact that they were found in shell marl is favorable to the idea that they belonged to the Pleistocene. Here may be mentioned again the bison horn of uncertain geological age which was found on the north shore of Nipissing Lake (p. 266). In Algoma County, on the banks of Moose River, was found a part of a skull of a mastodon, but there is uncertainty whether it had been buried in interglacial deposits or in marine Champlain beds. The region in the extreme eastern end of Ontario is interesting because it furnishes a considerable fauna belonging to the Champlain stage. During the last glacial stage the region on which the Wisconsin ice-sheet was resting became depressed to such an extent that when this ice retreated beyond the St. Lawrence River, marine waters occupied the basin nearly to the eastern end of Lake Ontario and Ottawa River as far as Lake Coulonge. Coleman’s figure of the region (Bull. Geol. Soc. Amer., vol. XII, pp. 129–146, fig. 1) is here reproduced (fig. 4) to show the western limits of the marine waters, so far as known, and the corresponding fresh-water beach along the north shore of Lake Ontario. Figure 5 from Coleman shows how the Champlain Sea was limited on the south. Marine fossils, especially mollusks, have been found along the upper St. Lawrence as far as Brockville, Quebec, and on the opposite side of the river, in New York. On Coleman’s map the present elevations of the old beaches at important localities are marked, that at Ottawa having an elevation of 450 feet and at Coulonge 370 feet. According to Johnston, who has described the Pleistocene geology in the vicinity of Ottawa (Mem. 101, Canad. Dept. Mines, 1917), there is a point about 8 miles northwest of the city where a marine terrace is found at a height of 690 feet above sea-level. The marine beds at Ottawa are divided into the Leda clays at the base and Saxicava sands above. The former have a maximum thickness of about 200 feet, the Saxicava sands, a thickness of about 40 feet. The fossils occur mostly in the Leda clays. In 1897, Dr. H. M. Ami (Ottawa Naturalist, vol. XI, pp. 20–26), and again in 1901 (Geol. Surv. Ann. Rep., XII, G, pp. 51–56), published lists of the fossils found in the Ottawa Valley, nearly all of them in the vicinity of Ottawa. There were listed 26 species of plants, about 13 species of marine mollusks, and the following vertebrates:

Mallotus villosus, capelin. Cyclopterus lumpus, lump-sucker. Osmerus mordax, smelt. Artediellus atlanticus (Cottus uncinatus), sculpin. Gasterosteus aculeatus, stickleback. Phoca vitulina, common seal (p. 22). Phoca grœnlandica, Greenland seal (p. 23). Tamias striatus, chipmunk.

The aquatic forms are all species existing in the Gulf of St. Lawrence and along the northern Atlantic coast. The chipmunk lives at Ottawa. Specimens of feathers of birds also have been found in nodules, but the species have not been determined. The remains of the chipmunk were probably washed in by some fresh-water stream.

According to Johnston’s paper just cited, there are deposits of glacial drift underlying the marine Champlain beds, but they have furnished no fossils. The marine deposits extend up the Ottawa Valley at least as far as Coulonge Lake, and here has been found _Mallotus villosus_. At Welshe’s, 3 miles north of Smith’s Falls, Lanark County, have been found some remains of the humpback whale, _Megaptera boöps_ (Dawson, Amer. Jour. Sci., vol. XXV, 1883, p. 200). It was met with (p. 17) at an elevation of 440 feet above present sea-level. It appears to have been left there during the time when the Saxacava sands and gravels were being laid down (Coleman, Bull. Geol. Soc. Amer., vol. XII, p. 133).

QUEBEC.

The Pleistocene of Quebec was described by Logan in 1863 (Geol. Canada, pp. 917–926) and by J. W. Dawson, 1894, in his “Canadian Ice Age.” Dawson divided the epoch, as represented in Canada, into the early Pleistocene, the mid-Pleistocene, and the later Pleistocene. He did not accept the glacial theory as it is now understood, admitting only great local glaciers. His early Pleistocene deposits embraced the great bulk of the boulder clays. His mid-Pleistocene represents an interglacial period, during which were deposited the marine Leda clays, Saxicava sands, and their fresh-water equivalents. The climate was supposed to be milder than at present. During the later Pleistocene there was to some extent a recurrence of local glaciation and of deposition of boulder clay. This stage was followed, according to Dawson, by the Early Modern, which he regarded as the age of the mammoth and mastodon.

Mr. J. Stansfield has described with some detail the Pleistocene and Recent deposits of the island of Montreal (Mem. 73, Geol. Surv. Canada, 1915). The boulder clay is of variable thickness and does not appear to be divisible into beds of different epochs. The Leda and Saxicava deposits are present. When the latter were laid down the region about Montreal was depressed about 600 feet below its present elevation. This has been confirmed by Goldthwait (Summary Rep. for 1913, p. 211). Later it began to rise; and Stansfield thinks that when the elevation had reached about 100 feet less than that of the present the water of the St. Lawrence at that point had become fresh. He found some apparent evidences of a recurrence of glaciation after the Champlain stage, but, on the whole, left the question undecided. He published a list of about 85 species of marine invertebrate fossils, collected from the Leda clay about Montreal, and 22 species obtained from the Saxicava sands. Besides the invertebrates secured from the Leda clays at that place, there are two vertebrates, _Phoca grœnlandica_ (p. 22) and _Delphinapterus leucas_, or _D. vermontana_ (p. 18). At Rivière du Loup, in Temiscouata County, whale remains were reported in 1894 (p. 18), which were thought to belong to _Delphinapterus leucas_. At Metis, Rimouski County, a jawbone of a whale has been discovered in the shelly marl of the lower terrace (p. 19); whether or not it belonged to _Megaptera boöps_ is not certain. The specimen of the former species was described by Leidy in 1856.

According to Logan’s report of 1863 (Geol. Canada, p. 920), the single bone was found in a brickyard. At the same place was found some vertebræ of the whale. At Bic, Rimouski County, has been found a nearly complete skeleton of a walrus, at an elevation of more than 100 feet (p. 21). Dawson (Canadian Record Sci., 1895, vol. VI, p. 352) described a nearly complete skeleton of the whale which had been found at Montreal in the Leda clay, 22 feet below the surface. This Leda clay was supposed by Dawson to have been deposited at a depth of from 50 to 80 fathoms, which depth, he said, corresponded approximately to the marine shore-lines at Montreal at an elevation of about 470 feet above sea-level, and to the sea-beach at Smith’s Falls, above referred to. Hence at the time that the whale was buried the mountain at Montreal was only a rocky islet in the sea which prevailed then over the region from the Laurentian hills on the north to the highlands of Quebec, south of the St. Lawrence.

At Tétreauville, in Ottawa County, on Ottawa River, have been found some bones, supposed to belong to the harbor seal, _Phoca vitulina_.

NEW BRUNSWICK, NOVA SCOTIA, AND CAPE BRETON ISLAND.

All three of these regions were involved in the glaciation of the Wisconsin stage. According to Goldthwait (Summary Rep. for 1913, pp. 244–250), New Brunswick was the center from which the ice flowed out over the other two lands. From this center it moved southward over the western end of Nova Scotia, more and more southeastward over the rest of the peninsula, while over Cape Breton Island the direction was eastward and northeastward. Some indications were observed of an earlier glaciation. As regards post-glacial submergence, Goldthwait found that at St. John, New Brunswick, this had amounted to about 190 feet, while on Cape Breton Island no signs of any submergence were found. Robert Chalmers had arrived at similar conclusions; and these agree well with the theoretical isobases drawn by Taylor for that region (Monogr. U. S. Geol. Surv. LIII, 1915, p. 503). G. F. Matthew in 1879 (Geol. Surv. Canada, Rep. for 1877–78, EE, pp. 1–36) described the geology of southern New Brunswick. Few fossil vertebrates of Pleistocene age have been discovered in these countries. On Cape Breton Island mastodon remains have been found in two places, Middle River and Baddeck (p. 46). As long ago as 1874 remains supposed to belong to _Delphinapterus_ were found near the mouth of the Jaquet River, in the northernmost part of New Brunswick; but Professor G. H. Perkins has shown that the animal was probably the narwhal, _Monodon monoceros_. The discovery is discussed here on page 19. At the southern extremity of New Brunswick, along Mace’s Bay, Charlotte County, a jaw supposed to belong to a species of _Delphinapterus_ was found, which had been buried in the Leda clay (p. 19). Near Fairville, at the mouth of St. John River, there has been discovered some bones of the seal _Phoca grœnlandica_ (p. 21). In the Academy of Sciences at Philadelphia is a skull of a walrus (p. 21) found apparently in the water near Sable Island about 50 years ago. It is not certain that it is a Pleistocene fossil.

NEW ENGLAND.

Inasmuch as relatively few vertebrates belonging to the Pleistocene have been discovered in the New England States, it will not be necessary to enter into details regarding the geology of the glacial period in this region. Nevertheless, the subject is one of great interest and one which has engaged the attention of many geologists. For those who wish to enter on the study, the writer recommends first a paper written in 1906 (Bull. Geol. Soc. Amer., vol. XVIII, pp. 505–556) by Frederick G. Clapp, entitled “Complexity of the Glacial Period in Northeastern New England,” which gives a brief history of the development of the idea that in the region mentioned there are evidences of more than one glacial and of more than one interglacial stage. There are also citations of the principal papers written on the subject. Among the writers cited are Shaler, Woodworth, Fuller, Upham, Stone, and Tarr. Clapp concluded that New England had been invaded by at least three ice-sheets and that these invasions had been separated by two interglacial intervals of long duration. On account of the greater thickness of the drift and because of fewer favorable exposures, due to the rocky nature of the coast and other causes, many difficulties are encountered in studying the deposits. He regarded absolute correlations as not yet possible. The last glaciation he accepted as corresponding closely with the Wisconsin, as displayed in States further west. What is known as Montauk drift, forming a part of the Gay Head interval of Woodworth, appeared to Clapp to correspond possibly to the Illinoian. Still older drifts would seem to have their place nearer the pre-Kansan (Nebraskan) than to the Kansan. What have been called “Leda clays” are found from Boston north into the St. Lawrence Valley. Clapp divides them into the “high-level” and the “low-level” clays. The former are the older and regarded as being about the equivalent to the Iowan stage. The “low-level clays” are referred to the Wisconsin stage. Another body of clays named by Fuller (Bull. Geol. Soc. Amer., vol. XVI, p. 375) the Gardiner clays, from their type locality, Gardiner Island, near the east end of Long Island, lies beneath the Montauk till and has been referred by Fuller to the Yarmouth interglacial.

In his paper cited Clapp presents (pp. 520–523) a list of the fossils, mostly mollusks, which have been collected in the Pleistocene deposits from New Brunswick to New York.

Along the New England coast are evidences of uplift which followed the retirement of the Wisconsin ice. Katz (Jour. Washington Acad. Sci., vol. VIII, 1918, p. 410) reported elevations of 155 feet at Stratham, New Hampshire, and 300 feet at Pawnal, Maine. Fairchild (Bull. Geol. Soc. Amer., vol. XXIX, p. 214) records the elevations at various localities in Maine.

A brief interesting account of the Pleistocene epoch as recorded in Massachusetts and Rhode Island may be found in an article by B. K. Emerson (Bull. U. S. Geol. Surv., No. 597, pp. 134–149). It deals in part with the geology of the valley of the Connecticut River.

Goldthwait (Appalachia, vol. XIII, pp. 1–23) and Foshay (Amer. Jour. Sci., ser. 4, vol. XXXVIII, pp. 345–348) have found evidences of an early Pleistocene glaciation in the White Mountains of New Hampshire.

Vermont is interesting especially on account of the Pleistocene history of Lake Champlain. This history has been recently discussed by Professor H. L. Fairchild (Rep. State Geologist Vermont, vol. X, 1916, pp. 1–41, with maps and views), who presents (pp. 40–411) a list, 37 in number, of the more important papers relating to the subject.

While the Wisconsin ice-sheet was resting upon Canada and the northern part of the United States, the land thus occupied, and probably a considerable area beyond the ice, became depressed. The valleys of the St. Lawrence, the Ottawa, the Hudson, and the Connecticut had been pressed down to such an extent that, as the ice-sheet retired these valleys became filled with water standing at sea-level. When at length the glacial front had retreated beyond the St. Lawrence, sea-water entered Lake Ontario and passed up Ottawa River far above the city of Ottawa (Leverett, Monogr. U. S. Geol. Surv., LIII, plate XXI). South of the St. Lawrence, marine waters occupied what is now Lake Champlain and as much of the surrounding land as was then at or below sea-level. In his account Fairchild makes use of the plate which is here reproduced (map 31) from his article of 1917 (Bull. Geol. Soc. Amer., vol. XXVIII, p. 279, plate XI). This geologist believes that the Hudson formed for a while a connection with Lake Champlain, although the Hudson waters may not have been actually saline. But in Lake Champlain the presence of fossil marine mollusks and at least one whale skeleton shows that its waters were salt. The lines crossing the plate obliquely are the isobases which show the amount of elevation which has taken place along those lines since the end of the Pleistocene. South of New York City this is zero. At the northern end of Lake Champlain the elevation is 800 feet. This means that the north end of the lake for a while stood 800 feet lower than now. Marine fossils have, however, been found at an elevation of only about 300 feet. The waters which first occupied the lake and stood at the highest level were of glacial origin and fresh. When the ice-front had receded so as to open the St. Lawrence and admit sea-water, the northern end of the lake had been uplifted about 500 feet. It was then that the marine animals entered.

Other important papers to be consulted in this connection are as follows: One by J. B. Woodworth (Bull. 84 New York State Mus.); one by Charles E. Peet (Jour. Geol., vol. XII, 1904, pp. 415–469; 617–660), and two by Professor Fairchild (Bulls. 105, 127, New York State Mus.).

It is proper to say that certain glacial geologists maintain that the depression in the New England States has been less than supposed by Fairchild, and that the isobases curved around toward the north as the New England coast was approached, somewhat as represented by Taylor (Monogr. LIII, U. S. Geol. Surv., p. 503). Fairchild, in a later paper (Bull. Geol. Soc. Amer., vol. XXIX, 1918, pp. 187–244), has reached the same conclusion and presented a map on which are drawn the isobases, or lines passing through points affected by the same amount of post-glacial uplift; from this map 32 has been prepared. On his map the location of the heavy or solid lines is regarded by Fairchild as being based on clear evidence. Where the lines become thin the evidence is less trustworthy; where the lines are broken their positions are hypothetical. The numerals on the lines show the amount of uplift along those lines. Two points of importance are brought out on the map. The first is that Newfoundland formed an independent center of glaciation and of subsequent uplift, a conclusion based on good geological evidence. The second point is that the center of the Wisconsin glaciation was located southeast of James Bay, considerably farther south and west than is usually supposed. The confirmation of this is left to the future.

It does not seem to have been demonstrated that there are in Connecticut any Pleistocene deposits older than those laid down by the Wisconsin ice-sheet. In case Fuller (U. S. Geol. Surv. Prof. Paper 82) is correct in his determination of beds of the early, middle, and late Pleistocene on Long Island, it is to be expected that beds of corresponding ages will yet be recognized in Connecticut. Woodworth (17th Ann. Rep. U. S. Geol. Surv., pt. 1, p. 978) mentions deposits of clay at Berlin and at New Haven that may be older than the Wisconsin.

While the correlations recorded above of the Pleistocene of the New England States with the glacial and interglacial stages of the Mississippi Valley may be subject to modifications, it is interesting to learn that the presence of Middle and Early Pleistocene deposits in the Eastern States has received the recognition of so many students of glacial geology. The hope is awakened that in New England there may yet be found interglacial deposits which will furnish remains of Pleistocene vertebrates, as these have come to light from Throg’s Neck, New York, to southern Florida. It is possible that the astragalus of an equine animal (p. 183), found at Gay Head, Martha’s Vineyard, belongs to a species of _Equus_ of early Pleistocene age.

In order to illustrate still further the events connected with the history of the Pleistocene in the region of the Great Lakes, three additional figures are introduced. One of these (map 33) shows J. W. Spencer’s conception of the drainage of the region in preglacial times. The areas now occupied by the lakes were then traversed by rivers. It will be observed that the rivers above Pittsburgh now discharging into the Ohio then emptied northward into the Erigan. This is shown also by a map (fig. 6) taken from Leverett (U. S. Geol. Surv. Monogr. XLI, p. 89). Figure 5, on page 287, shows the position of the shore of this Champlain Sea.

The number of Pleistocene vertebrates found in the New England States is limited, and most of them have been mentioned.

Somewhere on the coast of Maine have been found specimens of the fish _Mallotus villosus_ (Gould, Proc. Bost. Soc. Nat. Hist., vol. III, 1848, p. 67). At Charlotte, Vermont, a white whale, _Delphinapterus vermontanus_, was found many years ago (p. 19). Some bovid teeth were found many years ago at Gardiner, Maine, and referred to _Bison bison_, but it is now believed that they are teeth of the domestic ox. However, Dr. G. M. Allen has reported from Cape Cod, Massachusetts, teeth of a young bison (p. 266). At Woodbury, Washington County, Vermont, at a depth of 7 feet, an antler and a piece of the upper jaw with five molars of _Rangifer caribou_ (p. 244) have been discovered (Rep. Geol. Surv. Vermont, vol VI, p. 7). Mastodons have been discovered in Massachusetts at Coleraine and Shrewsbury (p. 47). Many years ago a tooth and a tusk and some bones of an elephant were found at Mount Holly, Vermont (p. 148); the writer refers the animal to _Elephas columbi_. An undetermined elephant has been found in Vermont at Richmond (p. 167). Walrus remains have been recovered at Addison Point (p. 23), Andrews Island (p. 23), Gardiner (p. 23), and Portland (p. 24), all in Maine; off Portsmouth, New Hampshire (p. 25); and on Martha’s Vineyard, Massachusetts (p. 25). At the latter place a tooth supposed to belong to the hooded seal (p. 26) was found long ago. With respect to the specimens found at this place there is some doubt as to their geological age. With the exception that the reindeer bones (p. 244) found near New Haven may be of pre-Wisconsin age, no Pleistocene vertebrate fossils older than Late Wisconsin appear to have been discovered anywhere in Connecticut. As shown elsewhere (p. 48), there were found long ago at Sharon, Litchfield County, remains which were identified as those of mammoth, but these have since been regarded as those of the common mastodon. Only a single vertebra was preserved.

Mastodons have been found in four other places, Cheshire, New Britain, Bristol, and Farmington (pp. 47, 48). The animals which left their bones at those places certainly lived after the last glacial sheet had withdrawn from the State. As mentioned on page 291, Fairchild has found reasons for believing that, while the Wisconsin ice-sheet was withdrawing from the Hudson and Connecticut Valleys, the whole region was so depressed that these valleys became occupied by water at sea-level. In these waters there were laid down thick deposits which now stand at levels much above tide, varying, in Connecticut, from nearly 200 to about 300 feet. Map 31, reproduced from Professor Fairchild (Bull. Geol. Soc. Amer., vol XXVIII, 1917, plate XI) is intended to show how wide an extent of territory along the Connecticut Valley was then submerged. It is probable that the emergence of these deposits was not accomplished until after the glacier had retired beyond the State.

It will be observed (map 6) that the localities just mentioned, where the mastodons have been found, lie very close to or on the areas covered by the deposits mentioned. The pond in which the Farmington mastodon (fig. 6, No. 3) was buried is in a range of hills which must have stood as an island in the Connecticut inlet. While it is possible that mastodons lived on this island while the land was depressed, it is more likely that they lived there after it had been more or less elevated. Judging from the topographical maps, one may conclude that the mastodons that have been found at Cheshire (fig. 6, No. 1) and New Britain (fig. 6, No. 2) were buried in deposits that overlie those laid down at sea-level. Their time of existence must have been near the end of the Pleistocene. Too little is known about the mastodons reported from Bristol and Sharon to form any definite opinion about the stage of the Pleistocene when they lived; but it was probably after the withdrawal of the last ice-sheet.

NEW YORK.

From the geologist’s point of view there is hardly, if at all, another State which presents for solution more numerous or more interesting problems connected with the Pleistocene than does New York. Among these are the geography and topography of the State at the beginning of the Pleistocene; the number and identity of the glacial stages which affected its surface; the origin and development of the bordering Great Lakes, of the numerous interior lakes, and of the river courses, actual and abandoned. For a knowledge of these one must consult the various reports issued by the Geological Survey of the State; above all, the numerous and instructive papers that have been published by Professor H. L. Fairchild, of the University of Rochester.

For the student of Pleistocene vertebrate palæontology, the State of New York is not so attractive as some others; but it is far from being devoid of interest. Few species of vertebrates of Pleistocene age have been found in its deposits, and these, with one exception, belong to the latest episodes of the last glacial stage. So far as the writer is aware, the following list comprises all of the Pleistocene vertebrates known to have been found within the borders of the State; those marked with an asterisk (*) are now extinct:

*Equus sp. indet (p. 183). *Platygonus compressus (p. 212) Bison bison (pp. 266, 267). Odocoileus virginianus (p. 226). Cervus canadensis (pp. 235, 236). Rangifer caribou? (p. 245). *Mammut americanum (pp. 48–63). *Elephas columbi (p. 149). *Elephas primigenius (pp. 131, 132). Castor canadensis. *Castoroides ohioensis (p. 272).

Deposits of materials belonging to Pleistocene stages older than the Wisconsin are apparently of rare occurrence in the State. If existing they are usually concealed beneath the widely spread Wisconsin drift. On Long Island, Fuller (U. S. Geol. Surv. Prof. Paper 82) has described beds of gravels, sands, and clays, which he regards as belonging to the Nebraskan, Aftonian, Yarmouth, and Illinoian. None of these has furnished any vertebrate fossils. However, in 1866 (Smithson. Contrib. Knowl., vol. XV, art. 3, p. 16), Whittlesey reported that he had a tooth of a horse (p. 183) found at Fort Schuyler, Throg’s Neck, 18 feet below the surface. This must have been lying beneath the Wisconsin drift. Inasmuch as Fuller has found the Manhassett formation, regarded as equivalent to the Illinoian, around Manhassett Bay, within 4 or 5 miles of Throg’s Neck, it seems entirely reasonable to suppose that deposits of similar or earlier age exist at Throg’s Neck.

With the exception of small areas, the whole of the State was at one time covered by the ice-sheet of the Wisconsin stage. The glacial ice filled the basins of the Great Lakes, and overrode even the peaks of the Adirondack and Catskill Mountains. Only along the southern side of Long Island and in the loop formed in Cattaraugus County by Allegheny River does the ice-sheet appear to have been absent.

Nearly everywhere, even on the southern coast of Long Island as outwash, it left its burden of clay, sand, gravel and boulders, usually many feet in thickness; in the mountainous regions this drift material is present, at least in the valleys. At the extreme southern edge of the glacial sheet there was laid down the terminal moraine, which, more or less distinctly determinable, has been traced from the eastern end of Long Island to the southwestern corner of Cattaraugus County, and onward into Pennsylvania. This moraine is shown here on maps 3 and 6–A.

As the ice-sheet withdrew toward the north, the surface which it had occupied was, for many reasons, very uneven, and in the depressions there were formed numerous lakelets and lakes. Into the smaller lakes and ponds especially, were swept, by running water and blown by winds, coarse materials and dust, so that they began at once to fill. Water-loving plants in due time took possession of their borders, and in time marshes were formed. In some of these bodies of waters are now found deposits of shell marl, which show that for a long period the lakes and ponds were inhabited by fresh-water mollusks. Sometimes below this marl, but usually above it, is found a layer of peat, the product of the partial decay of the vegetation. It is in such peat-bogs, sometimes buried in the peat, sometimes in the marl, that have been found most of the bones and teeth of the fossil animals recovered. Inasmuch as such deposits lie upon the Wisconsin drift, it is certain that these animals lived, at the localities where found, after the retirement of the glacier from that locality; how long afterward one usually can not be certain.

It is in such Late Wisconsin deposits that have been found the numerous remains of mastodons on Long Island, on Staten Island, around New York City, and especially in Orange County (pp. 48–54). This county has furnished some of the most complete skeletons of mastodons ever discovered. Whether or not the conditions for their existence were more favorable in this region than in that between this county and the Finger Lake region may be regarded as doubtful; but it is certain that the conditions for the preservation of skeletons were extremely favorable.

A remarkable case is presented at Cohoes, where a part of a skeleton of a mastodon was found in one of the great pot-holes existing there, and another part of the same skeleton in a neighboring pot-hole. The case is discussed below.

In the western half of the State, after the foot of the glacier had retired beyond the divide between the present northward and southward flowing streams, bodies of water began to collect between the divide and the foot of the glacier. To these bodies, regarded as lakes, changing from time to time their dimensions and their outlets, have been given various names. At first, the waters that collected in the Finger Lake region found their outlet southward through the Susquehanna River; later through the Mohawk and Hudson; then westward into the Mississippi drainage; afterward through a channel leading around west and north of the Adirondacks and into Lake Champlain and down the Hudson; and finally, as now, into the St. Lawrence River (map 34).

The waters of the Erie basin, for most of the time, found their outlet toward the west into the Mississippi; but at a later time they escaped for a while eastward through central New York into the Mohawk. For information regarding these lakes one must consult Leverett and Taylor (Monogr. LIII, U. S. Geol. Surv.) and Fairchild (Bulls. 127, 160, N. Y. State Mus.).

From a study of the geological history we may arrive at some approximately correct ideas as to the time when the mastodons, elephants, horses, giant beavers, etc., lived within the limits of the State. Of these animals, apparently none of the specimens discovered up to this time belongs to any pre-Wisconsin stage, except the horse whose tooth was found at Throg’s Neck (p. 183). The history of our extinct horses and the depth at which the specimen was found indicate that the animal had lived either during the first or the second third of the Pleistocene.

We may be certain that none of the mastodons (p. 49) which have been reported from Long Island lived there while the northern border was occupied by the glacier, and the remainder by the ocean. Not until the land had risen to about its present level could mastodons have become buried in the muck-filled ponds where they have been met with. Where the glacier front was when mastodons got foothold on the island we can not tell certainly; but it required perhaps hundreds or probably thousands of years for the elevation of the island to the extent of about 100 feet. We can hardly doubt that the mastodon lived on up to near, possibly into, the Recent period (see map 34).

It is interesting to speculate on the time and manner of entombment of the skeleton, described on page 56, which was found at Cohoes, part in one pot-hole, part in another not far away. Hall adopted the theory that the carcass of the mastodon had been frozen in the glacial ice and, on the thawing of this ice, had been dropped into the pot-holes. In fact, he thus explained the frequent presence of mastodon skeletons in swamps. We have, however, no evidence that mastodons were ever thus frozen up in the ice of the glacier; but there is a possibility that this happened sometimes. If a skeleton should thus have been engaged in the moving stream of ice it is not probable that it would ever have emerged in a recognizable condition. In the production of cracks and crevices in the glacial ice, of which Hall spoke, the bones would have been broken up and scattered, if not ground to powder. If a cadaver had been frozen in the ice for any considerable time it would certainly have come out in such a waterlogged condition that it would hardly have floated. Weighted down by its heavy tusks, it would have drifted against rocks and at least the tusks would probably have been broken off. If we exclude the idea that the mastodon had first been frozen in the glacier, the writer sees no reason for denying that it might thus have been transported for some distance; but little is gained by granting it. The animal could as well have lived near Cohoes as farther up the Mohawk.

As stated on another page, James Hall concluded that the pot-holes belonged to some preglacial time. Professor H. L. Fairchild has expressed in a letter to the present writer the following opinion:

“When the ice-sheet melted from Cohoes the locality was 355 feet lower than it is to-day. Deep estuary deposits partially filled the Hudson Valley and buried the Cohoes district. The Mohawk channel at Cohoes is excavated through marine sediments. There is no suggestion of any river channel there previous to the present river work. The pot-holes are post-glacial, but they probably represent a more copious and vigorous flow than that of the present river. That was supplied by the diminishing Iromohawk, the latest outflow through the Mohawk Valley of the Iroquois water. In this view the pot-holes were drilled by the latest glacial waters.”

It appears that, when the mastodon skeleton fell into the pot-holes, these had been drilled long before; for the principal one had become filled with gravel to a depth of at least 10 feet. They were, therefore, probably well above the stream-level, except in times of high-water. However the carcass reached the locality, it must have arrived in a complete state. Had it already attained an advanced stage of decay, some limbs or the feet or the lower jaw, probably the whole head, weighted down as it was by the heavy tusks, would have dropped off. It may be assumed that the skeleton was lying on land or in some pond not far above the pot-holes. The flesh was not wholly decayed, and the bones were held together by the ligaments. While the skeleton was in this condition the river rose and swept it over the first pot-hole, where the right leg dropped off; and then onward over the second, where more of it was deposited. Some unimportant parts may have been carried farther, and some of the missing bones may have decayed in the pot-holes. After the bones were deposited there the pot-holes became slowly filled up, probably mostly during times of high-water, with muck and branches and trunks of trees of several species (Hay, Science, n. s., vol. XLIX, 1919, p. 378).

The retreat of the Wisconsin ice-sheet far beyond the St. Lawrence and the rise of the land to its present elevation, 350 feet above the sea at Cohoes, belong to the closing chapter of Pleistocene history. When the Cohoes mastodon was buried the ice-sheet was probably already north of the St. Lawrence and, as Professor Fairchild writes, 150 feet of the rise of the land had already occurred. The time could, therefore, not have been long before the beginning of the Recent epoch. If these animals lived at such a late time at Cohoes they doubtless existed at the same time in all parts of the eastern region where their remains have been discovered. They may have been able to occupy Long Island a little earlier than places further north, but the interval would be geologically inconsiderable.

The writer has learned of no discoveries of mastodon bones in materials laid down by the marine waters that occupied Lake Champlain, the St. Lawrence Valley, and that of Ottawa River, or in deposits overlying these marine beds.

On the basis of one of Professor H. L. Fairchild’s plates (Bull. 127, N. Y. State Mus., plate XXXV) the writer has prepared map 34, which is intended to show the position of the Wisconsin ice-sheet in New York after it had retired somewhat north of the divide. This divide is marked by a line of dots. The area then occupied by the ice is stippled. Lake Erie was already nearly free from ice and was discharging its water by way of the Mississippi. Impounded waters from the melting glacial ice were collecting in the region of the Finger Lakes, forming Newberry Lake, and escaping down the Susquehanna. The Mohawk afforded outlet for the water from the southeastern lobe of ice. Fairchild’s plates 36 to 42 show the successive positions occupied by the ice-front as it retired northward and the various lakes that were formed.

Although not many species of vertebrate animals have been found in the Pleistocene deposits of New York, a large number of localities have furnished remains of the mastodon, _Mammut americanum_. These localities are recorded and brief descriptions of the remains and their geological environment have been presented on pages 48–63. The localities are indicated on map 34. It will be seen that several specimens have been found on Long Island and many in Orange County, in the southeastern corner of the State. In the western half of the State most of the finds occur within the area once occupied by the successive lakes. The animals could have lived there only after the ice-sheet and the lake waters had disappeared. It will be seen that a few finds have been made close to the shores of the present lakes. The animals must have lived there at the very end of the Pleistocene, if not within the Recent epoch.

The finds of other vertebrates are recorded on the following pages: _Equus_ sp. indet. on page 183; _Platygonus compressus_ on page 212; _Bison bison_ on page 266; _Odocoileus virginianus_ on page 226; _Cervus canadensis_ on page 235; _Rangifer caribou_ on page 245; _Elephas columbi_ on page 149; _Elephas primigenius_ on page 131; _Castor canadensis_ on page 272; _Castoroides ohioensis_ on page 272.

In 1850 (Proc. Amer. Assoc. Adv. Sci., vol. II, pp. 255–256), W. C. Redfield reported that he had received remains of a fox of the genus _Vulpes_ from Gulf Summit, Broome County. The lower jaw and other bones had been discovered in a cutting of the New York and Erie Railroad, 40 feet below the natural surface. The deposit above these bones was evidently the Wisconsin drift. The fine clay inclosing the bones may have belonged to the Sangamon, or even some older interglacial deposit. It is impossible to say whether this fox was _Vulpes fulvus_ or _Urocyon cinereoargenteus_.

NEW JERSEY.

(Map 6–A.)

In the consideration of the problems of Pleistocene geology and palæontology, New Jersey is one of the most important States. Its northern part is occupied by glacial drift deposits, while the southern two-thirds is covered more or less completely by materials laid down beyond the limits of the glaciers. The glacial materials appear to belong to two widely separate epochs. The geologists who have been connected with the geological survey of New Jersey recognize in the materials composing the Pleistocene deposits south of the glacial region three formations, the Bridgeton, oldest; succeeded by the Pensauken; and the Cape May, the youngest. The geologists of Maryland recognize in New Jersey three formations which correspond to the three of Maryland, the Sunderland, the Wicomico, and the Talbot. However, the author of the Maryland Pliocene and Pleistocene volume, Professor Shattuck, insists that parts of Salisbury’s Bridgeton, Pensauken, and Cape May all enter into the Sunderland; parts of the Cape May, Pensauken, and possibly of the Bridgeton, into the Wicomico; and parts of the Pensauken and Cape May into the Talbot.

There are wide divergences in the views of the two groups of geologists regarding the manner in which the materials have been laid down. The Maryland geologists hold that their three terraces represent three epochs of submergence, and that the gravels, sands, and clays were deposited in the salt waters of the ocean or of estuaries. Salisbury and Knapp (Geol. Surv. New Jersey, vol. VIII, 1917, p. 3) adopt the view that the formations are partly of subaerial and partly of marine and estuarine origin, with emphasis on the subaerial mode. Of the Bridgeton, the authors referred to say (their p. 18 ) that the accessible parts are primarily of terrestrial origin. A part of what remains may be marine or estuarine, and part of what has been removed may have been so. No palæontological evidences of marine deposits of this epoch are found in the State. The writer records his dissent from the theory that the terraces and the deposits called the Sunderland, Wicomico, and Talbot have been the product of marine submergence. A part only of the Talbot can be referred to deposition in the sea.

Of the Pensauken, Salisbury and Knapp say (p. 87): “There is nothing in its constitution to negative the hypothesis of the whole formation being river work; nor is there anything, as now understood, to prove it.” As to the deposits which they refer to the Cape May, the authors quoted say (p. 162) that the southern part of the State seems to have stood a few feet (30 to 50) lower than at present; but that it could not have stood long at this height, for sea-cliffs are essentially wanting. At one point, near Millville, Cumberland County, marine fossils are met with at an elevation of about 10 feet above tide.

The Cape May was, according to Salisbury and Knapp, laid down during the last glacial epoch, the Wisconsin (p. 162). This determination of age would doubtless gain the acceptance of the Maryland geologists and their adherents, although the latter would include under this name many local deposits which Salisbury puts in the Pensauken.

It is remarkable that, so far as the writer knows, no remains of Pleistocene vertebrates have ever been discovered in that portion of New Jersey which is mapped as occupied by the Cohansey sands, an area including nearly half the State. It lies southeast of a straight line which would run from Navesink River to Salem. The reason for this lack of fossil vertebrates does not occur to the writer. A large portion of this region is mapped as being covered with deposits of all three of the Pleistocene formations, Bridgeton, Pensauken, and Cape May. On or near to the line of outcrop of the Cretaceous deposits from Salem to Raritan Bay, not fewer than ten localities are known where mastodon remains have been discovered, besides two localities which have furnished horses and two which have furnished elephants. Since the southeastern part of the State has yielded no vertebrate fossils and little else to throw light on the age of its deposits, we shall dismiss it from consideration.

The glacial geology of the State has been studied by Professor Rollin D. Salisbury, of the University of Chicago, and his assistants, Henry B. Kümmel, Charles E. Peet, and George N. Knapp. The results of their studies on the glacial-drift deposits have been published in volume v of the final report of the State geologist, 1902.

The Quaternary formations of the southern part of the State are described in volume VIII of the final report. A more succinct description of the events of the Quaternary period is found in Bulletin 14 (1915) of the New Jersey Survey. The authors are J. Volney Lewis and Henry B. Kümmel.

In the vicinity of Perth Amboy is a heavy glacial moraine which may be traced eastward through Staten and Long Islands. West of Perth Amboy it turns northward, and swinging around it reaches Springfield. Thence it runs northwestward to Rockaway, and continues west by south to Delaware River, at Belvidere. This moraine marks, in New Jersey, the southward limit of the last ice-sheet, the Wisconsin. All the drift deposits of the State north of this moraine are regarded as belonging to the Wisconsin stage. It is to be supposed that this is, at least to some extent, underlain by older drift deposits.

South of the moraine just described are scattered deposits of glacial drift and other evidence of glacial action which are referred to a much older ice-sheet, one supposed to correspond to the Kansan drift of the Mississippi Valley (Salisbury, Geol. Surv. New Jersey, vol. V, p. 781). On the other hand, it is sometimes referred (Chamberlin and Salisbury, Geology, vol. III, pp. 383, 384) to the first glacial (sub-Aftonian).

As has been said, three formations are recognized which were laid down otherwise than by glacial ice-sheets, the Bridgeton, the Pensauken, and the Cape May. The deposition of the Cape May is regarded as being contemporaneous with the Wisconsin ice-sheet (Salisbury and Knapp, New Jersey Geol. Surv., vol. VIII, p. 162; Lewis and Kümmel, Bull. 14, p. 120). The Pensauken formation is believed to be much older than the Cape May; it may (Salisbury and Knapp, op. cit., p. 78) be older than the extra-morainic drift, mentioned above as being of about Kansan times; but it may have coincided in part only with the Kansan. According to Lewis and Kümmel (op. cit., p. 111) the old, extra-morainic, Jerseyan drift was coincident with at least the later stages of the Pensauken. Hence, we may believe that the Pensauken corresponds somewhat to the Aftonian stage of Iowa. The Bridgeton formation is still older than the Pensauken and, being Quaternary, must be referred either to the early part of the first interglacial or to the first glacial; but the New Jersey geologists are not specific on this point.

It is unfortunate that nowhere in New Jersey has any considerable number of species of Pleistocene vertebrates been found buried together. We are thus deprived of one means of estimating the age of the species and of the beds. Most of the specimens found, as the mastodon and the two elephants, belong to species which lived during the whole or a large part of the Pleistocene and hence do not testify definitely to the age of the deposits in which they occur. Too often the information we have regarding the place and conditions of burial is extremely meager.

In Salem County a mastodon has been found in Mannington Township, at Chestnut Hill (p. 63); and a deer, probably _Odocoileus virginianus_, at Woodstown (p. 226). Although the geological map shows that in Mannington Township Cape May Pleistocene prevails, while about Woodstown there is Pensauken, one can not well conclude that the animals are of corresponding age.

In Gloucester County _Mammut americanum_ has been found at Harrisonville (p. 63), Mullica Hill (p. 64), and Woodbury (p. 64); _Equus_ at Swedesboro (p. 184). As to the former species, we can not be certain of the age, either from our knowledge of the age of the deposits inclosing the remains or from the history of the species. As to the horse found at Swedesboro, one may, from the history of the genus in this country, arrive at some conclusion; but this will be deferred to page 303.

In Camden County, so far as the writer has knowledge, no vertebrate remains have been found except in the Fish House beds, along Delaware River, just above Camden; but the horse remains (p. 184) are of great importance. These beds were originally supposed to be of Cretaceous age, but in 1869 (Trans. Amer. Philos. Soc., vol. XIV, p. 250), Cope expressed the conviction that they belonged to the Pliocene period. He presented a geological section (fig. 7) of the beds which shows a thin stratum of soil above, then from 8 to 15 feet of light-brown sand, followed below by a blackish clay about 25 feet in thickness. Near the bottom of the latter was found a layer containing shells of several species of _Unio_ and _Anodonta_. Just above this bed of unios there was discovered a large part of a skull of an extinct horse which Cope referred to _Equus fraternus_. This was deposited in the collection of the Academy at Philadelphia, but later disappeared.

In 1897 (Rep. State Geologist, New Jersey, for 1896, pp. 201–247, plates X-XIV) Woolman published a paper on the stratigraphy of the Fish House beds and described and illustrated other horse-teeth which he referred to _Equus complicatus_. These teeth were found at a depth of 12 feet below the top of the black clay; 6 feet of surface gravels had been removed from the clay. The teeth are now in the collection of the Academy, at Philadelphia. Woolman stated that in the same collection are a patella and a fragment of a long bone of a horse found in the black clay, in 1892.

Woolman regarded the clay in question as belonging to Pensauken times. Salisbury and Knapp (op. cit., p. 104, fig. 49) state that there is here 20 feet of black clay overlying Pensauken sand and that the clay is overlain by Pensauken gravel. If this judgment of the geological age of the clay is correct, the horses probably lived during the first interglacial (Aftonian) or the beginning of the second glacial stage (Kansan). There are, however, those who insist that these Fish House clays belong really to the Cape May formation. This would make the geological age of the horse about that of the Wisconsin drift.

Besides the horse remains, only some bones of a wolf have been found in the clays mentioned, and these too have disappeared. They probably would have thrown little light on the age of the beds. We must reach conclusions from other data.

This fact seems to be pretty certain: Had horses lived at Fish House during the deposition of the Cape May they would (as did the mastodon, _Elephas primigenius_, and _E. columbi_) quite certainly have spread out over northern New Jersey and over the grassy plains of New York and Ohio; and their remains would somewhere have been found, as are those of the other species just mentioned, in old swamp and lake deposits overlying the Wisconsin drift; but no horse remains have ever been reported from such deposits. Furthermore, in all the digging that has been done at Trenton, in deposits acknowledged by all to belong to Wisconsin times, no trace has been found of horse remains.

Near the bottom of the Fish House clay bed, just below the level of the horse remains, there is found a layer which contains river clams represented by the genera _Unio_ and _Anodonta_. Ten species of _Unio_ have been recognized and two of _Anodonta_. When these were first studied the beds were believed to belong to the Cretaceous. Nevertheless, the close resemblance of the shells to still living species was recognized; and to them were given names differing from those of the related existing forms by the ending _oides_. The species were described by Lea and Whitfield and have been restudied by Dr. H. A. Pilsbry, of the Academy of Natural Sciences, Philadelphia. The species are probably identical with forms yet living; but half of them no longer exist in the region of Delaware River. Pilsbry (Proc. Acad. Nat. Sci. Phila., 1896, pp. 567–570) stated that five of them have no longer any representatives in the Atlantic drainage south of the St. Lawrence River system. It is probable that these species had, when they lived at Fish House, spread into other rivers south of the Delaware and thus were not trapped in this river by the Wisconsin ice. It seems certain, therefore, that a longer period of time and a longer series of vicissitudes must have intervened to produce such changes in geographical distribution. According to C. T. Simpson’s work, “Descriptive Catalogue of the Naiades,” 1914, _Unio (Quadrula) subrotundus_ now inhabits the Ohio, Cumberland, and Tennessee Rivers; _U. (Lampsilis) anodontoides_ occupies the Mississippi River and Gulf drainage regions; while _Anodonta corpulenta_ is found in the Upper Missouri region. The Wisconsin ice-sheet and the short period of time since its disappearance are hardly sufficient to explain this wide dispersion of species, while others have been able to retain their place in the Delaware.

Opposed to this view regarding the identity of the unios of the Fish House beds, see Ortmann (Proc. Amer. Philos. Soc., vol LII, p. 280, 1913) and Baker (Univ. Ills. Bull., XVII, p. 205, 1920). These writers contend that the species have no especial relationship to western forms. According to Baker the deposits are older than the earliest glacial stage. On the other hand, according to Dr. E. W. Berry (quoted by Baker), who has studied the plants, the beds belong to the late Pleistocene.

We have, then, these reasons for holding that the Fish House clays are of early Pleistocene age: (1) Competent geologists have determined them as belonging to the Pensauken formation, laid down at or before the time of the Kansan stage; (2) the presence of remains of horses, evidences of whose existence during or after the Wisconsin have not been produced; (3) the presence of many species of naiades, some of which yet live in that region, but the majority of which now live only in far-distant regions.

We may confidently conclude that the horse remains which were found at Swedesboro belonged likewise to the Pensauken.

In Burlington County mastodons have been found at Pemberton ( p. 64), but one can not be certain of their geological age. A reindeer has been unearthed at Vincentown (p. 64). It seems highly probable that it lived there while the Wisconsin ice-sheet occupied the northern part of the State; but there is a possibility that it is older. In the Academy of Natural Sciences at Philadelphia are some remains of _Odocoileus_ found at Vincentown (p. 227).

In the vicinity of Trenton, Mercer County, scant remains of six species of Pleistocene mammals have been reported. These are _Mammut americanum_ (p. 64), _Elephas primigenius_ (p. 132), Bison bison (p. 287), _Ovibos moschatus_ (p. 248), _Cervus canadensis_ (p. 237), and _Rangifer caribou?_ (p. 248). All are known to have existed elsewhere during late Pleistocene times, and three indicate a cold climate. The presence of fossil vertebrates here is of special interest because many evidences have been found of man’s occupation of the region in apparently late Pleistocene times.

At and in the vicinity of Trenton are found both Pensauken and Cape May deposits, the latter overlying the former (Salisbury and Knapp, op. cit., pp. 120, 165). The Cape May rises about 60 feet above sea-level. At various places the Pensauken protrudes through the mantle of Cape May and rises to a height of as much as 130 feet above sea-level. Its base is about 20 feet above sea-level. The materials consist of sand, gravel, and cobblestones. So far as the writer knows, no fossils have been found in the Pensauken about Trenton.

The Cape May at Trenton is held to have been laid down principally during the presence of the Wisconsin ice-sheet in the northern part of the State; and naturally it consists mostly of sands, gravels, coarse and fine, and some boulders. In the localities where excavations have been made for sand and gravel for building purposes, for sewers, and for railroads, and in search for relics of man, two principal divisions are recognized. Below are strata of clays, sands, gravels, and boulders which are believed to have been deposited by the floods of varying intensity which issued from the glacial moraine then about 60 miles above Trenton (figs. 8, 9). Over this lies a bed of what is called yellow drift, which reaches a thickness of about 3 feet. It consists mostly of fine sand, but there are many pebbles and occasionally some large boulders. It is everywhere characterized by wavy red bands. While some geologists have held the opinion that this deposit had been produced by winds, it appears to be definitely determined that it was waterlain (Wissler, Scient. Monthly, vol. II, p. 237). This “yellow drift” is overlain by about a foot of black soil which belongs to the Recent epoch and is the result of cultivation by whites. For details regarding the Trenton gravels and the yellow sands above it the reader should consult Ernest Volk’s work, “Archæology of the Delaware Valley” (Papers Peabody Mus., vol. V, 1911).

All the species mentioned above have been reported from the beds known as the Trenton gravels. A femur of a bison was found also in the yellow drift (see p. 287).

Monmouth County has furnished more fossil vertebrates of the Pleistocene than any other county. Mastodons have been discovered at Englishtown, Freehold, Marlboro, Long Branch, Manasquan, and in the Navesink Hills (pp. 65, 66). Many specimens, as those about Freehold and Long Branch and Manasquan, are in such superficial positions in peat that they do not seem to be very old, probably of Cape May age; and yet of this one can not be wholly certain. The discovery of a heel-bone of a megatherium (p. 31) at Long Branch appears to indicate the presence there of early Pleistocene deposits. At Englishtown the remains had apparently become mixed with marl, and they may belong to an older stage of the Pleistocene. In the Navesink Hills, according to Leidy, the mastodon remains were associated with those of an extinct horse (p. 184). If so, both species probably were buried in Pensauken deposits. In this same region there was found long ago a tooth of _Elephas columbi_ (p. 149); but it is useless to speculate on its geological age. At Long Branch (p. 26), damaged skulls of walruses, probably of the existing species, have been met with. It seems natural to associate this southward migration, which extends to South Carolina, with the Wisconsin epoch; but it is possible that it was earlier. At Deal (p. 227) have been found remains of a deer, probably _Odocoileus virginianus_.

Somewhere about Shark River, a tooth of a peccary (p. 213) was found, as was supposed, in Miocene marl. Leidy could not distinguish this tooth from that of _Mylohyus nasutus_. So far as our evidence goes, this species belongs to the early and middle Pleistocene.

Near North Plainfield a tooth was found which is referred to _Elephas primigenius_ (p. 133). The locality is very close to the moraine of the Wisconsin ice-sheet, and the animal probably lived there when the Plainfield outwash plain (Salisbury, Geol. Surv. New Jersey, vol. V, 1902, p. 738) was being laid down.

Near Schooley’s Mountain, but west of Musconetcong River and in Warren County, remains of a mastodon (p. 67) were encountered in excavating the Morris Canal. It is probable that these were buried in a swamp left over from the Wisconsin times; but Lewis and Kümmel’s map of 1910–1912 indicate in this region only drift older than the Wisconsin.

The mastodons found at Hackettstown and Hope, in Warren County, are probably of Late Wisconsin origin (pp. 67, 68).

Near Mount Hermon, about 5 miles northeast from Delaware, in Warren County, and about 2 miles northwest of Hope, was found the splendid skeleton of the moose _Cervalces scotti_, which forms one of the treasures of Princeton University (Scott, Proc. Acad. Nat. Sci. Phila., 1885, p. 174). It was discovered in a bog. All this region is (Salisbury, Geol. Surv. New Jersey, vol. VIII, plate XXVIII) occupied by Wisconsin drift and the bog doubtless rested on this drift. It seems certain, therefore, that this stately relative of our existing moose lived after the disappearance of the Wisconsin ice-sheet.

A mastodon (p. 68) which was found at Greendell in Sussex County quite certainly lived there after the last glacial stage.

Berry (Torreya, vol. X, p. 261) has studied a collection of nine species of plants which had been obtained in peat from near Long Branch. Only three of these now range north of Long Branch. He concluded that the last glacial stage had been followed by a period of climate warmer than the present climate. This is in accord with views which the present writer has held. It ought not, however, to be assumed with too much confidence that the peat-bed is of Late Wisconsin origin.

PENNSYLVANIA.

About half of the area of Pennsylvania lies outside of the region which was glaciated. Figure 10 is a map taken from Folio 172 of the U. S. Geological Survey, published in 1910 and compiled by Dr. W. C. Alden in 1901. A broad strip along the southern part of the State, being non-glaciated, is not represented. The areas shaded by parallel ruling and stippling are those which present evidences of glacial action.

The glaciated area consists of two principal portions. One of these, that subjected to the action of the Wisconsin ice-sheet, is represented on the map by means of oblique parallel lines coming down to an interrupted heavy line. This line, representing the Wisconsin terminal moraine, starts on Delaware River north of Easton, runs northwestward to Potter County, thence into New York, thence back into Pennsylvania, in Warren County, and then enters Ohio north of the Ohio River. The course of this moraine was worked out especially by H. C. Lewis and G. F. Wright and was described in report L of the Pennsylvania Geological Survey, in 1881. The moraine crosses the Delaware at Belvidere, New Jersey, and passes through the following counties: Northampton, Monroe, Carbon, Luzerne, Columbia, Sullivan, Lycoming, Tioga, Potter, Warren, Crawford, Venango, Butler, Lawrence, and Beaver.

South of this moraine are two areas which, on this map, are represented by stippling. These are occupied by drift materials, usually forming a considerably thinner covering, which are believed by most glaciologists to belong to an older Pleistocene stage, probably about as old as the Kansan. Especially in the valleys these older drift deposits may reach thicknesses of 200 or 300 feet. These old glacial deposits are represented also by terraces along the margins of the valleys. Some of these in the vicinity of Warren stand at a height of about 1,400 feet above the sea. Figure 17 is taken from Shaw and Munn (Folio 178, U. S. Geol. Surv., p. 12). The uppermost gravels are supposed to represent the Kansan stage. A few small patches lying in the angle of the unglaciated area are of doubtful age, as indicated on the map. It must be stated, however, that there is some dissent from this conclusion as to the age of this outer drift. Professor E. H. Williams has published a number of papers in which he takes the position that this drift is a deposit laid down by the same ice-sheet that later on built up the Wisconsin moraine (Amer. Jour. Sci., vol. XLVII, 1894, pp. 32–36; Science (n. s.), vol. XXXVII, pp. 447–450; Pennsylvanian Glaciation, first phase, 1917, pp. 1–101). Professor G. F. Wright appears to take the same view. The writer sees no sufficient reason for distrusting the opinions of Dr. Alden and his colleagues.

It must not be assumed that an animal whose remains have been found within the area occupied by the Wisconsin drift lived during or after that stage. Even within this area there may occur fossil-bearing deposits of an older Pleistocene time. These older deposits may underlie the Wisconsin drift or they may occur as old terraces high up on the sides of the valleys of rivers. Cases of the latter kind are found along Allegheny River (Leverett. Monogr. XLI, U. S. Geol. Surv., pp. 229–252; Shaw Jour. Geol., vol. XIX, 1911, pp. 140–156; folio 178, U. S. Geol. Surv., p. 8). On the other hand, an animal of very late Pleistocene age, or even of the Recent, may be buried in deposits which overlie an old Pleistocene deposit. It is necessary, if it can be done, to determine the actual age of the deposit containing the remains; otherwise one must depend on the geological age of the species involved, or be content to wait for further information. Unfortunately, but few of the quadrangles in the glaciated area have had their geological structure studied and reported on. At present the U. S. Geological Survey has published only Folios 92 (Gaines) and 93 (Elkland and Tioga), lying mostly in Tioga County, partly in Potter; also Folio 172 (Warren), occupying a part of Warren County. Information may sometimes be secured from the numerous volumes which have been published by the Geological Survey of Pennsylvania and from articles in the scientific journals.

The Pleistocene deposits which lie outside of the glaciated areas have been mostly laid down along rivers. Some of the materials were transported by the streams which carried away the drainage from the glaciers; in other cases the materials were brought down from the higher lands and laid down along the lower and less sloping parts of the streams. In the unglaciated area many of the quadrangles have been surveyed by the U. S. Geological Survey and the folios aid in determining the age of deposits which contain fossil vertebrates.

Important collections have been made in a few localities, and these will now be considered:

At Pittston, in Luzerne County, on Susquehanna River, have been found teeth of the horse _Equus complicatus_ (p. 184), remains of mastodon (p. 68), and of a musk-ox (p. 248). The presence of the horse makes it evident that the deposit containing the fossils belongs to a stage older than the Wisconsin, although the locality is within the area of the Wisconsin.

We consider now the contents of a cave found near Stroudsburg, Monroe County. The Hartman (or Crystal Hill) Cave was discovered in 1880 and explored first by Mr. T. Dunkin Paret, of Stroudsburg. It was soon afterward examined by Dr. Joseph Leidy, of Philadelphia, and Dr. Thomas C. Porter, of Easton. Leidy published the first description of it in 1880 (Proc. Acad. Nat. Sci. Phila., pp. 346–348) and presented a list of the species of animals which had been secured by Mr. Paret. In 1889 (Ann. Rep. Geol. Surv. Pennsylvania, for 1887, pp. 1–18, plates I, II), a more detailed report was made by Leidy, including descriptions and illustrations of some of the vertebrates and of certain artifacts which had been discovered.

In 1894, Dr. H. C. Mercer made a re-exploration of the cave and gave a more extended description of it (Proc. Acad. Nat. Sci. Phila., pp. 96–104).

Combining the statements of Leidy and Mercer with data obtained from the Delaware Water Gap topographical sheet issued by the U. S. Geological Survey, one finds that the cave is situated on Crystal Hill, about 3.5 miles in a straight line southwest of Stroudsburg and close to the village of Stormville. Crystal Hill is a part of an anticlinal fold, Godfrey Ridge, of the Helderberg limestone. South of the fold runs Cherry Creek; north of it, Mt. Michaels Creek. On the northeast the hill is cut off from the rest of the ridge by a valley about 300 feet deep. Mercer’s account states that the cave is on the top of the hill, about 0.25 mile from Cherry Creek, but the topographical map locates the top of the hill about 0.75 mile away from this stream. Mercer also wrote that the cave was 800 feet above Delaware River, 5 miles away. However, the hill has an elevation of somewhat less than 840 feet above sea-level, while the river at the nearest point is somewhat more than 280 feet above sea-level. Inasmuch as the cave is probably somewhere on the southern slope of the hill, it is about 500 feet above the Delaware and about 300 feet above the bed of Cherry Creek.

The opening of the cave in the rock was wide (Mercer, p. 96, fig. 1), but had become almost wholly choked by débris. Nevertheless, a hole large enough for adventurous boys to enter remained (Leidy, op. cit., 1880, p. 346). After a few feet descent the cave extended nearly horizontally more than 100 feet. It had become filled nearly to the roof by various deposits. Excavations showed that on top was a layer, about a foot, of “black friable earth mingled with animal and vegetal remains” (Leidy). Mercer describes it as a “top layer of limestone roof-splinters and down-slidden outer talus thinning inward into less stony cave earth.” Beneath this layer was a thin stratum of stalagmite. Further digging showed that below this stalagmite flooring the cave was filled to a thickness of as much as 14 feet in one place. This deposit is described by Mercer as being a continuous homogeneous bed of exquisitely fine clay deposited in thin laminæ rarely sprinkled with sand pockets and underlain with a thin film of sand. Neither in this deposit nor in the stalagmite was there found a trace of any formerly living thing. All the remains of animals and all the artifacts were discovered in the uppermost layer.

It should be noted at this point that this cave is situated about 5 or 6 miles north of the Wisconsin moraine.

The following is a list of the species of vertebrates identified by Leidy. When his names differ from those now in use they are inclosed in parenthesis.

_List of species of vertebrates._

Chelydra serpentina. Terrapene carolina (Cistudo clausa). Meleagris gallopavo sylvestris (M. gallopavo). Equus sp. indet. (p. 185). Mylohyus pennsylvanicus (Dicotyles) (p. 213). Rangifer caribou (p. 246). Odocoileus virginianus (Cervus) (p. 237). Cervus canadensis (p. 237). Bison bison? (B. americanus) (p. 267). Marmota monax (Arctomys). Tamias striatus. Sciurus carolinensis. Castor canadensis (C. fiber). Peromyscus leucopus (Hesperomys). Neotoma magister (N. floridana). Microtus pennsylvanicus (Arvicola riparius). Erethizon dorsatum. Castoroides ohioensis (p. 272). Sylvilagus floridanus (Lepus sylvaticus) Myotis subulatus (Vespertilio). Eptesicus fuscus (Vespertilio). Scalopus aquaticus (Scalops). Procyon lotor. Mustela noveboracensis (Putorius ermineus). Mephitis putida (M. mephitica). Urocyon cinereoargenteus (Canis virginianus). Canis lycaon? (C. lupus). Lynx canadensis (Felis).

Besides these vertebrates, there were reported by Leidy the land snails _Helix albolabris_, _H. alternata_, and _H. tridentata_; also a pair of valves of the river mussel _Margaratina margaritifera_ and a fragment of another valve. Leidy regarded these as showing that this mussel formerly lived in Delaware River; whereas in his view it no longer existed there; but specimens of it from Philadelphia are in the U. S. National Museum.

An examination of the list shows that nearly all of the species of vertebrates are yet in existence and most of these still living in that general region. _Rangifer caribou_ lives now far to the north and _Lynx canadensis_ has its range somewhat further north. The two indicate a colder climate, especially the reindeer. Both got into the cave probably after the glacial front had withdrawn from that vicinity. The remains of _Castoroides_ may have been carried in there at about the same time. The type specimen of _Mylohyus pennsylvanicus_ was found in this cave. Cope referred specimens of a peccary found in Port Kennedy Cave to the same species with doubt. Undetermined species of the genus were recognized by Barnum Brown in his collection made in the Conard fissure in northwestern Arkansas. Dr. W. J. Holland reported _Mylohyus pennsylvanicus_ from the cave at Frankstown, Pennsylvania. The type of the genus, _M. nasutus_, was found in Indiana. Beyond the testimony furnished by the Crystal Hill Cave, we have no evidence that the genus _Mylohyus_ existed after the Wisconsin stage; the possibility exists that this species got into the cave before this stage.

The specimen of _Equus_ is still more doubtful. It consisted of two isolated first and second milk molars of a very young colt. Leidy was in doubt whether the colt belonged to the domestic horse or to an indigenous species. The specimen had been collected with no record as to the part of the cave or of the depth in the upper layer of soil where it was buried. A fragment of a jaw of a colt might easily have been carried into the cave by some carnivorous animal since the coming of the whites. A fragment of the lower jaw of a bison also was found which had in it the last molar; and this was referred by Leidy to the existing buffalo.

It can hardly be doubted that this cave was hollowed out before the Wisconsin ice period. It may have been formed during the early Pleistocene. The fact that it was filled to a depth of 14 feet, in some places, with a fine laminated clay devoid of all traces of organic beings seems to indicate that for ages it had been shut off from the outer world, and that streams charged with fine sediment were permitted to pass through it. During possibly some glacial stage preceding the Wisconsin, erosion may have opened the cave so that the horse remains, those of a bison, and of _Castoroides_ were dragged into it. The evidence for these suppositions is slender, but so too is that for a late Wisconsin indigenous species of horse in Pennsylvania. It is probable that most of the species found in the cave belong to the late Pleistocene or even to the Recent.

Fossil vertebrates found in a cave in Bucks County require our attention.

In 1880 (Proc. Acad. Nat. Sci. Phila., 1880, p. 349), Leidy presented a list of vertebrate remains which had been lying unstudied for 40 years in the collection of the Academy. These had been found in Durham Cave, somewhere near Riegelsville, in Bucks County. It is not improbable that the cave took its name from the village of Durham, about 2 miles southwest of Riegelsville. Leidy stated that the cave appeared to have been obliterated in the quarrying of limestone. In 1889 (Ann. Rep. Geol. Surv. Pennsylvania, for 1887, pp. 18–19) Leidy published a list of the species which he had identified.

_List of fossil vertebrates from Durham Cave._

Acipenser sturio. Ameiurus nebulosus (Amiurus atrarius). Thamnophis sirtalis (Eutænia). Chelydra serpentina. Terrapene carolina (Cistudo clausa). Meleagris gallopavo sylvestris (M. gallopavo). Rangifer caribou (p. 246). Cervus canadensis (p. 237). Alces americanus (Alce). Odocoileus virginianus (Cervus) (p. 227). Erethizon dorsatum. Marmota monax (Arctomys). Sciurus carolinensis. Castor canadensis (C. fiber). Neotoma pennsylvanica (N. floridana). Ondatra zibethica (Fiber). Sylvilagus floridanus (Lepus sylvaticus). Ursus americanus. Procyon lotor. Mephitis putida (M. mephitica). Urocyon cinereoargenteus (Vulpes virginianus).

This list differs in its species from Leidy’s list of 1880 only in the exclusion of the bison and the inclusion of the elk, _Cervus canadensis_. All the species are still in existence, most of them in that region. The presence of the reindeer, the moose, and the porcupine suggests a cooler climate than now prevails there. These animals may all have become buried in that cave during the latest times of the Wisconsin stage or even during the Recent.

We are now to study a case which furnishes us with a store of knowledge regarding the life of the Pleistocene. In 1871 there was found at Port Kennedy, Montgomery County, a cave which was worked for its fossils by Charles Wheatley and later by Dixon, Mercer, and Cope, the latter having devoted himself to the description of the vertebrates. First of all will be given a list of the species of vertebrates, mostly mammals which have been recognized in the materials found in the cave. When Cope’s names differ from those employed here they are put in parenthesis.

_List of species of vertebrates found in Port Kennedy Cave._

Ranidæ: Rana sp. indet. Emydidæ: Clemmys insculpta. C. percrassa. Terrapene eurypygia (Toxaspis anguillulatus). Colubridæ: Coluber acuminatus (Zamenis). Meleagridæ: Meleagris superbus (M. altus). Megatheriidæ: Megalonyx (p. 31). M. scalper (p. 31). M. tortulus (p. 31). M. wheatleyi (p. 31). Mylodon harlani? (p. 31). Equidæ: Equus complicatus (E. fraternus) (p. 185). E. pectinatus (E. f. pectinatus) (p. 185). Tapiridæ: Tapirus haysii (p. 203). Tagassuidæ: Mylohyus nasutus (p. 213). M. pennsylvanicus? (p. 213). Tagassu tetragonus (Mylohyus) (p. 213). Camelidæ?: Teleopternus orientalis (p. 224). Cervidæ: Odocoileus lævicornis. O. virginianus? Bovidæ: Bison sp. indet. (Bos) (p. 256). Elephantidæ: Mammut americanum (Mastodon) (p. 69). Sciuridæ: Sciurus calycinus. Castoridæ: Castor canadensis (C. fiber). Cricetidæ: Peromyscus leucopus? (Hesperomys). Anaptogonia hiatidens. Sycium cloacinum. Microtus dideltus. M. diluvianus. M. involutus. M. speothen. Zapodidæ: Zapus hudsonius? Erethizontidæ: Erethizon dorsatum? Ochotonidæ: Ochotona palatina (Lagomys). Leporidæ: Sylvilagus floridanus (Lepus sylvaticus). Talpidæ: Scalopus sp. indet. (Scalops). Soricidæ: Blarina simplicidens. Vespertilionidæ: Myotis? sp. indet. (Vespertilio). Ursidæ: Ursus americanus. Arctotherium haplodon. Mustelidæ: Taxidea taxus (T. americana). Mephitis fossidens. M. leptops. M. obtusata. M. orthostica. M. putida. Osmotherium spelæum. Pelycictis lobulatus. Mustela diluviana. Gulo luscus. Lutra rhoadsii. Canidæ: Canis priscolatrans. C. dirus? (C. indianensis). Urocyon cinereoargenteus. U. latidentatus. Felidæ: Machairodus gracilis. Smilodontopsis mercerii (Smilodon). Felis eyra. F. inexpectata (Uncia). Lynx calcaratus.

Into this list there are admitted 60 species, of which 54 are mammals. Of these, 41 are extinct, not counting the doubtful species unless there is good reason for it. There are, therefore, 68 per cent of the species extinct.

No remains of _Rana_ were mentioned by Cope in his list of 1899. One species unnamed was recorded by Wheatley in his lists of 1871 and by Mercer in his paper of 1899. The turkey (_Meleagris superbus_) was not included by Cope in 1899, but it was included by Wheatley and Mercer and Cope in their papers of 1871 and in that of Cope in 1896 (p. 378). Mercer (1899, p. 280) mentions a leg-bone of a turkey, with spur, found by Wheatley. Remains of _Megalonyx_ were abundant, but of _M. loxodon_ only a single tooth was met with. _Mylodon_, believed to be _M. harlani_, was found only by Wheatley and was represented, as stated by Cope, by only a claw phalanx. The horse remains were originally (Cope, 1895, p. 447) referred to _Equus major_ (=_E. complicatus_). Mercer, in 1899, in his figure 9, following Cope’s nomenclature, uses the name _E. complicatus_. In 1899, Cope concluded that the equine remains represented two races of _Equus fraternus_, _E. f. fraternus_ and _E. f. pectinatus_. The present writer believes that the teeth referred to the subspecies _fraternus_ are too large to belong to the species which was called _E. fraternus_, but which is now called _E. leidyi_. Only a single species of tapir, _Tapirus haysii_, was recognized. Cope (1895, p. 447) stated that it was the most abundant of the larger mammals. Cope (1899, p. 257) reported that 18 individual peccaries were represented by teeth, while bones were numerous. He recognized the presence of three species. The identifications of _Mylohyus nasutus_ and _M. pennsylvanicus_ were uncertain. A new species, _M. tetragonus_, was based on a ramus of a lower jaw. Milk molars were yet present and the third molar had not appeared. Cope spoke of the long diastema; but, to judge from his figure, the diastema equals only about the length of the milk molars and the first molar.

Cope, in 1899, described _Teleopternus orientalis_, basing it on a few teeth which belonged to three individuals. He was doubtful about the family position of the animal, but put it provisionally in the Camelidæ. In many respects the teeth resembled those of the Cervidæ. Matthew (Osborn, Age of Mammals, p. 469) has suggested its affinity to _Ovibos_.

Two species of deer were found in the cave, of which one was not distinguishable from _Odocoileus virginianus_. In Wheatley’s second list of 1871 and that of Cope of the same year there was recorded an undetermined species of _Bos_ (_Bison_). Mercer (1899, p. 280) recorded from the Wheatley collection remains of three individuals of one species of the same genus. In Cope’s paper on the remains of this cave nothing is said about the genus; but in 1872 (Proc. Amer. Philos. Soc., XII, p. 96) he stated that _Bos_ was represented by a part of a femur and some other bones. Hence in the list given above an undetermined species of _Bison_ is included.

Abundant remains of the mastodon occurred in the cave, but none of any of the elephants. One need not, however, on that account conclude that elephants were not living in that region at that time.

It will be observed that a considerable number of rodents is included in the list. One species of porcupine is recognized. This was at first regarded by Cope as an extinct form and called _Erethizon cloacinum_; but in 1899 he referred all the remains, with some doubt, to the existing species, _E. dorsatum_. Cope found remains of about 50 individuals of a species of rabbit which he determined as _Lepus sylvaticus_, but this is now called _Sylvilagus floridanus_. In the Wheatley collection a species of bat was recognized and put in _Vespertilio_. Probably it belonged to _Myotis_.

Bears were abundantly present in the cave. One species, _Arctotherium haplodon_, was larger than the grizzly bear and represented by parts of about 25 individuals. A smaller bear, indicated by 8 individuals, appeared to be in no way different from the existing black bear, _Ursus americanus_. Of skunks there are listed 7 species, belonging to 3 genera, all the species being extinct except a supposed _Mephitis putida_. Besides these mustelids, there have been identified remains of the existing badger, the existing glutton, an extinct weasel, _Mustela diluviana_, and an extinct otter, _Lutra rhoadsii_. Remains of true dogs were not abundant in the collection. Cope recognized, however, 2 species of the genus _Canis_, one of about the size of the more common form of the existing wolf; the other exceeding in size the largest wolf known to him. This he thought might belong to Leidy’s _Canis indianensis_ (=_C. dirus_ Leidy). There were present 2 foxes, the existing gray fox (_Urocyon cinereoargenteus_) and an extinct species, _U. latidentatus_. Of the cat family a species, thought at first to be a hyæna (_Crocuta_), received the name _Felis inexpectata_. It had the size of the jaguar, and was represented by teeth and various bones. An extinct lynx, much like _Lynx ruffus_, was present. Another cat was identified as _Felis eyra_. Of this species G. S. Miller (Bull. 79, U. S. Nat. Mus., p. 116) remarks that its type locality is Paraguay and that it is supposed to range north to Central America. It appears somewhat doubtful, therefore, that the fossil remains belong to this species. Nevertheless, the progenitors of the species, in their wandering from Asia or Alaska to Central America and Paraguay, might have sent a colony into Pennsylvania, later to become extinct. Cope stated (1899, p. 250) that there was an isolated calcaneum in the collection which was of the proper size for _Felis eyra_, but which differed from that of this species. Two species of saber-tooth cats were found, _Smilodontopsis gracilis_ and _S. mercerii_. The former is represented by various bones and teeth, especially by a damaged skull which presents the dentition. The crown of the great canine is 113 mm. long.

Besides the species included in the list given above, there are a few whose presence for one reason or another is doubtful. In both of his lists of 1871 Wheatley reported the presence of _Crotalus_, _Coluber_, and _Tropidonotus (Natrix)_. Cope (1871, p. 98) said that the reptiles included three or four serpents, but in 1895 (p. 447) he wrote that two species of _Ophidia_ were recognized. In his final paper he mentioned only his _Zamenis acuminatus_, here referred to _Coluber_. Wheatley (1871, p. 255) recorded an unidentified snipe as belonging to _Scolopax_. Cope (1871, p. 98) wrote that a snipe was one of two species of birds present. Mercer (1889, p. 280) recognized the same remains as belonging to a species of _Gallinago_. Wheatley in his last list (1871, p. 384) and Cope (1871, p. 98) reported _Scalopus (Scalopus)_ as being represented by an undetermined species. It is catalogued by Mercer in the same way. Cope (1895, p. 447) stated that the raccoon was very rare; but it was not mentioned in any of his later papers. On the same page he wrote that there were fragments of teeth closely similar to those of _Bassariscus astutus_; but the species was not mentioned afterward.

As already said, there are admitted into the list given above, as identified in a reasonably good manner, 60 species, of which 54 belong to the Mammalia. It is a matter of interest to compare these with the species of mammals which were living in that general region before the fauna was disturbed by the arrival of the whites. The number of species of the existing mammals, as shown in the second column, is obtained from Rhoads’s “Mammals of Pennsylvania and New Jersey.” The subspecies are not included.

_Families of land mammals represented in Port Kennedy Cave and those that have lived in that region within Recent times, together with the number of known species in each family at each of the two epochs._

┌──────────────────────┬──────────────────────┬──────────────────────┐ │ Families. │ No. of species, Port │ │ │ │ Kennedy. │No. of recent species.│ ├──────────────────────┼──────────────────────┼──────────────────────┤ │Megatheriidæ │ 5│ │ │Didelphidæ │ │ 1│ │Equidæ │ 2│ │ │Tapiridæ │ 1│ │ │Tagassuidæ │ 3│ │ │Camelidæ? │ 1│ │ │Cervidæ │ 2│ 2│ │Bovidæ │ 1│ 1│ │Elephantidæ │ 1│ │ │Sciuridæ │ 1│ 6│ │Castoridæ │ 1│ 1│ │Cricetidæ │ 7│ 9│ │Zapodidæ │ 1│ 2│ │Erethizontidæ │ 1│ 1│ │Ochotonidæ │ 1│ │ │Leporidæ │ 1│ 2│ │Soricidæ │ 1│ 5│ │Talpidæ │ 1│ 3│ │Vespertilionidæ │ 1│ 8│ │Procyonidæ │ ?│ 1│ │Ursidæ │ 2│ 1│ │Mustelidæ │ 11│ 9│ │Canidæ │ 4│ 3│ │Felidæ │ 5│ 3│ └──────────────────────┴──────────────────────┴──────────────────────┘

In the column of fossils there are 54 species; in that of the Recent there are 58 species. Of two families represented at present in the region, but not included in the Pleistocene column, Didelphidæ and Procyonidæ, the latter named has had remains referred to it with doubt. Without doubt members of both families existed there at that time.

Of the families of the Pleistocene column two no longer live anywhere near the region; four nowhere on the continent; one nowhere on the earth. Even of such families as the Ursidæ and the Felidæ important elements, as _Arctotherium_ and the saber-tooths, are extinct. Of the 54 species admitted in the Pleistocene column 40 are extinct; that is, 74 per cent.

If we consider the sizes of the animals in question we gain this result: Only 15 of the existing species are of any considerable size, ranging from that of a raccoon to that of a bison, about 26 per cent. Of the 54 fossil species of mammals, about 30 vary in size as indicated, about 57 per cent. It is hardly to be doubted that this preponderance is due to the poorer chances which the smaller skeletons had of preservation and of rescue from the matrix. Had the smaller fossil species been preserved and collected in the same proportion that the smaller existing ones have to the larger, the cave ought to have furnished twice as many species of mammals as it did. It is, of course, possible that the larger species are more liable than the smaller ones to become extinct as time passes on. We can hardly doubt, in any case, that when the Port Kennedy animals were being buried in that cave there lived in that region a considerably larger number of species than within Recent times. There must have existed in that region more moles, more rabbits, more cricetids, more squirrels, and many more bats. Certainly there is no adequate record of the number of birds, snakes, turtles, and amphibians that must have existed about Port Kennedy and have perished in that cave.

From the collection that has been made in the cave at Port Kennedy some definite conclusions ought to be reached regarding their time of existence. In his account of the cave and of the exhumation of the animal and vegetable remains, Mercer (1899, pp. 269–286) has shown what extreme care was taken in recording the position which each specimen occupied in the deposits. In his figure 9 he has noted the levels which the various species occupied. While the existence of four beds of materials makes it evident that the deposition went on for some time, it is noted that few or no differences exist in the character of the species included. Possibly Mercer’s subdivision 1 is to be excepted in this statement. Certainly no great changes went on in the fauna while the cave was being filled; no such changes as occurred in the glaciated region from the Aftonian interglacial stage up to the Late Wisconsin. It appears more probable that the deposits in the cave and the animals entombed there appertain to about a single Pleistocene stage. Is, then, the stage the Late Wisconsin?

This cave is situated only about 55 miles south of the Wisconsin moraine. At the time the species found in the cave existed they must each have occupied a wide extent of territory. It is not to be doubted that the range of nearly every species extended northward far beyond the moraine mentioned. Why, then, in deposits overlying the Wisconsin drift have there never been found any remains of the four Port Kennedy species of _Megalonyx_, of _Mylodon_, of the two species of horses, of the tapir, of the three species of peccaries, of the deer _Odocoileus lævicornis_, of the five extinct species of cricetids, of _Ochotona_, of the extinct species of _Blarina_, of the great bear _Arctotherium_, of the six extinct species of skunks, of the extinct otter, of the extinct dog, of the extinct fox, of any species of saber-tooth tiger, or of the extinct cats _Felis inexpectata_ and _Lynx calcaratus_? The absence of so many species of animals, most of them of large size, from deposits so well adapted to preserve bones and teeth, render it very certain that the animals no longer existed there.

Did the extinct species which are referred to above exist in eastern Pennsylvania at some time during the Wisconsin glacial stage and perish before the close?

A few of the species found in the cave and still existing are at present inhabitants of regions somewhat more northerly than Port Kennedy. Such are _Erethizon dorsatum_ and _Gulo luscus_; but the great majority, living and extinct, indicate a climate at least as warm as that of the present; many of them suggesting a still milder condition. Within historical times both of the species just named have inhabited the Alleghany Mountains at least as far south as Port Kennedy. Cope, in 1871 (Proc. Amer. Philos. Soc., vol. XII, p. 99), concluded that he had then identified in the cave remains of 11 neotropical species. It appears, therefore, wholly improbable that this assemblage of animals lived in that region, so close to the foot of the glacier, during the Wisconsin stage. These animals must have had their time of existence previous to this inhospitable epoch. It seems to the writer that the proportion of extinct species, three-fourths, and the history of many of the genera and species, indicate a time about equivalent to the Aftonian.

Professor A. Heilprin (Proc. Phila. Acad. Nat. Sci., 1895, p. 451) expressed himself as being inclined to refer this cave fauna to the Pliocene. An examination of this opinion would show that it is no more tenable than the opinion that the fauna is of the Wisconsin stage. It will not be discussed here beyond saying that deposits containing a similar fauna are found along the Atlantic coast from New Jersey to the Gulf, and that at one place at least, Vero, Florida, these are underlain by abundant Pleistocene sea-shells.

Besides the vertebrates which have been listed, a number of beetles were found and about 10 specifically determined plants. Wheatley (1871, p. 385) presents a list of the beetles as determined by Dr. G. H. Horn, but the names were not accompanied by descriptions. When later (Trans. Amer. Entom. Soc., vol. V, 1876, pp. 241–245) Horn came to describe them he reduced the number of species and, in some cases, gave them other names. The following is a list as given in Horn’s paper just cited: _Cychrus wheatleyi_, _C. (minor)_, _Pterostichus_ (spp. indet.) _Cymindis aurora_, _Chlænius punctulatus_, _Dicælus alutaceus_, _Choeridium? ebeninum_, _Phanæus antiquus_, _Aphodeus precursor_. All of these, as the writer is informed by Dr. E. A. Schwarz, of the U. S. Department of Agriculture, are regarded as extinct, but as closely allied to species now living in that general region. The plants, as reported by Mercer, are _Quercus palustris_, _Q. alba_, _Q. macrocarpa_, _Fagus ferruginea_, _Corylus americana_, _Pinus rigida_, _Carya porcina_, _C. alba_, _Ampelopsis quinquefolia_, _Cratægus crus-galli?_, and all still flourish in eastern Pennsylvania.

Mercer (1899, p. 269) has given a description of the cave found in quarrying operations. It was located on the right bank of the Schuylkill River, at the village of Port Kennedy and about 2 miles below Valley Forge. Wheatley (1871, p. 236) gave a map which showed the position of the quarries. A comparison of this with the topographical map of Folio 162 of the U. S. Geological Survey shows that they were situated about 800 feet away from the river and facing the valley of an unnamed streamlet. None of the descriptions give the elevation of the cave above the river or above the sea. The river at that place is apparently about 70 feet above sea-level. The 100–foot contour-line runs along near the location of the quarries, but these may have extended back to a higher level. Putting all of the statements together, it appears probable that the mouth of the cave was, in Wheatley’s time, about 50 feet above the level of the river. Originally the surface elevation may have been still greater, but may have been reduced by erosion of the hill. The surface rock here is red shale of the Stockton formation, belonging to the Triassic, and is underlain by the Shenandoah limestone, a member of the Cambro-Ordovician series. This limestone was being quarried in 1871, when a cave was broken into, filled with incoherent materials and exposing fossil bones in abundance. It was visited by Charles Wheatley, who proceeded to make excavations and collect the fossils. In studying the fossils he worked with Professor E. D. Cope and Dr. G. H. Horn. The results were published in Wheatley’s two papers of 1871 and in two papers by Cope in the same year (Proc. Amer. Philos. Soc., vol. XII, pp. 15, 73–102). According to Wheatley’s description and his figures, the part of the cave seen was about 20 feet wide at the top, expanded below to about 30 feet, and then narrowed at the bottom, as then recognized, to about 10 feet. The depth was given as 40 feet, but Mercer thinks that this was improbable and that Wheatley’s measurements were to some extent guesses. Mercer (1899, p. 271) stated that this cave might be compared to a bottle of unknown size. It had opened to the surface; and on his page 283 Mercer spoke of it as forming a well-like hole that might have been as much as 70 feet deep. Evidently Mercer here included that part of it which he himself excavated. The materials filling it were, according to Cope (1871, p. 73), the débris of the neighboring Triassic strata. Figure 11 is taken from Mercer’s paper and is a reproduction of a sketch made by Wheatley in 1871. After Wheatley had made his collection the cave was covered over by débris from the quarry and forgotten.

In the course of further quarrying operations the same cave was broken into again in 1893. Excavations in the materials that filled the cave were made in 1894 by Dr. Samuel Dixon, H. C. Mercer, and others, resulting in the securing of the collection which formed the subject of Cope’s paper of 1894 and his final report of 1899. At this time, according to Mercer, the quarrying operations carried on from 1855 had transformed a gently sloping hillside into an amphitheater several acres in extent, walled with perpendicular escarpments of rock, sometimes a hundred feet high. At this time the floor of the quarry had been lowered and the cave was broken into at a level below that reached by Wheatley. Figure 12, reproduced from Mercer’s figure 5, shows the relation of the later excavations to those of 1871. As already stated, Mercer concluded that Wheatley’s dimensions were probably results of guesses, inasmuch as the top of Mercer’s exposure was not more than 30 or 33 feet below the original level of the hilltop. According to Mercer’s figure 5, his own excavation probably extended down about 16 feet below the level reached by Wheatley; but other statements appear to make this somewhat greater.

Mercer wrote that the materials filling the cave had been stratified by the action of water. He recognized four subdivisions, most of which stood higher around the walls than at the center of the cave. Of these subdivisions, the first and uppermost was supposed to mark the lowest level attained by Wheatley. It consisted of fine clay and loam of black color, intermingled with fine and coarse muck, in which were found some remains of small mammals, just what species was not stated. On his chart, his figure 9, a tapir is indicated as occurring in it. Subdivision 2 was composed of from 4 to 11 feet of sandy clay, with fragments of sandstone and limestone, from small ones up to about 2 feet in diameter. In this matrix there were numerous bones and teeth of large animals, but it lacked small ones and vegetal matter. Subdivision 3 was a sandy clay, blackened by vegetable matter and containing numerous bones of vertebrates, large and small. The lowest subdivision, 4, was a zone which was followed down about 10 feet and which consisted of sand, clay, and stones, all of a yellow color. In this were found remains of the larger mammals, better preserved than in the upper subdivisions. At the lowest depth reached the excavation appears to have extended below the level of the Schuylkill River and the water came in so rapidly that further descent was not practicable.

Mercer’s theory of the filling of the cave is expressed in these words, on his page 277:

“Enough had been seen to convince us that a fresh-water flood, rising to a level of from 15 to 20 feet above the present level of the hilltop, hence a general inundation of the whole surrounding country, bearing in its current the clay, stones, and earth of neighboring levels, had tumbled into the fissure, carrying with it the bones of creatures previously denuded of flesh and softened by decomposition.”

And further, on page 284:

“Not unreasonably, therefore, we may suppose, not only that the creatures had perished together, but also that they had perished on the spot or at the chasm—not meeting this fate during a long interval of time, and through a long series of chance tumbles, but suddenly and by force of a common event.”

Are we to suppose that during some summer freshet animals in such numbers were swept away that those that were found in the cave, and doubtless many more which decayed utterly, were only the relatively few that happened to pass over that 20–foot hole? Where, then, were picked up all the other animals that must have burdened the swollen Schuylkill? Or did it possibly happen that all the animals that were swept away were in some unaccountable manner directed into that hole? If the current was strong enough to sweep along stones up to 2 feet in diameter, how did it happen to deposit there fine sand and clay, leaves, cones, seeds, and sticks? It is difficult to accept the theory that the filling of the cave was due to a cataclysm such as has been invoked. It seems far more probable that the mouth of the cave was open for many hundreds of years, possibly thousands of them, so that animals, plants, stones, and fine and coarse earth could in various ways get into it. Animals wandering about might inadvertently fall in or be pushed in by the herd. Doubtless at some former time the Schuylkill flowed at a higher level than now, and during times of unusually high-water might have risen to the level of the mouth of the cave and carried into it at each rise some mud, some vegetation, and some animals. The filling was quite certainly a slow process.

To the writer the part of the cavern which was worked and pictured by Wheatley has all the marks of an enormous pot-hole, such as those which have been discovered at Cohoes, New York. While the latter appear to have been drilled out in late Pleistocene times, the Port Kennedy hole must have been fashioned during the early Pleistocene or even in the Pliocene. One may suppose that, after the pot-hole had reached the depth where the constriction was found, the water began to find its way out at the bottom through fissures or passages in the limestone. When this happened, the passages may have been enlarged mechanically or by means of solution, resulting in the formation of the various lower caverns. When the river had been lowered enough to reach only occasionally the mouth of the pot-hole, the latter became choked first by the coarse materials now found in subdivision 4, and afterwards by finer sand and mud.

Some vertebrates of the late Pleistocene or early Recent observed at Carlisle deserve consideration.

In 1850 (Proc. Amer. Assoc. Adv. Sci., vol. II, 1849, pp. 352–355) Professor S. F. Baird gave an account of his explorations in the caves in the region about Carlisle, Cumberland County. One of these caves was near Carlisle, and in it Baird found a large number of animal remains. A second cave, the situation of which was not given, was on the top of a hill and was a vertical shaft 30 feet deep, which opened into a large gallery. It furnished a skeleton of a bear, but this appeared to have only recently fallen into the cave. Another cave was on the bank of the Susquehanna, 0.5 mile below a railroad bridge. It was, therefore, probably near Harrisburg. The entrance was in limestone rock, nearly vertical, and 20 feet deep. Here Baird found many bones, embedded in mud, but of these he obtained only a few. Another cave, apparently nearby, which Baird spoke of as “the main cave,” furnished some of his specimens. Still another cave, probably in the same neighborhood, was the source of his most perfect specimens. This presented a series of galleries near the roof and these were reached by ladders. These galleries were filled with mud, and in this mud the bones were buried. The number of species which he obtained, he reported, was nearly twice the number living there at the present time. Of these fossil species he estimated that about 5 per cent were extinct. Baird appears never to have completed his study of his collection. His list designates the animals only by their vernacular names. The mammals consisted of panthers, lynxes, wolves, foxes, otters, bears, muskrats, deer, beavers, and rabbits. There were bird remains in great quantities, and these included wild turkeys, some of great size, swans, wild ducks, and pelicans. There appeared to be 8 or 10 species of tortoises. Bones of snakes were quite common; also scales and vertebræ of fishes, and a lower jaw of a salamander. In the uppermost 2 or 3 inches of mud were many relics of Indians.

Baird supposed that these bones had in most cases been washed in from above through sink holes. This collection, or some of it, was brought by Baird to the Smithsonian Institution; and they, or some of them, are in the collection of mammals; but the bulk of the collection has apparently been lost. All of these animals belong evidently to either the very late Pleistocene or to the Recent period.

A cave at Frankstown has furnished fossils of about Middle Pleistocene time. In 1908 (Ann. Carnegie Mus., vol. iv, pp. 228–233) and again in 1912 (Proc. Internat. Zool. Congr., Boston, 1907, pp. 748–752), Dr. W. J. Holland gave an account of the discovery of vertebrate fossils in a fissure in limestone rock at Frankstown, Blair County. This village is situated on the Frankstown Branch of Juniata River, a little more than 2 miles north of east of Hollidaysburg. The fissure was excavated in a Devonian rock known as the Lewistown limestone. The quarries are reported to be in the village and on the top of a hill that rises about 400 feet above the banks of the Juniata. According to the Hollidaysburg topographical sheet, the 920–foot line crosses the river just above the village. The highest hill, 1,260 feet above sea-level, is 0.3 mile away toward the northwest. In this hill, as Dr. Holland stated, there are several small caves. The one which furnished the fossils appeared to be about 40 feet in length, averaging from 6 to 8 feet in width, and at the most was not more than 10 or 12 feet high. The floor was about 30 feet below the top of the hill. The fissure appeared to have once continued up to the surface, but the opening had been filled with fallen blocks of limestone. The floor of the cave is described as being occupied by about 2 feet of red soil, everywhere traversed by bands and layers of dark materials charged with organic matter. With the finer deposits were mingled fragments of rock, some being large blocks. The fossil remains appear to have been carefully collected, but were mostly fragmentary. They were only cursorily studied at the time of Holland’s writing and nothing has since been published on them. The number of species obtained was estimated to be from 30 to 40. The following genera and species are mentioned:

Meleagris sp. indet. Megalonyx sp. indet. (p. 31). Tapirus sp. indet. (p. 203). Mylohyus pennsylvanicus (p. 214). Odocoileus virginianus (p. 227). Cervalces? sp. indet. Bison sp. indet. Mammut americanum (p. 69). Sciurus sp. indet. Ondatra sp. indet. Erethizon sp. indet. Lepus sp. indet. Ursus americanus. Arctotherium haplodon. Mephitis sp. indet. Canis sp. indet. Felis? sp. indet.

After the foregoing had been put in type Mr. O. A. Peterson, of the Carnegie Museum, sent the writer a revised list in which additions are made. The following are the most important:

Cryptobranchus sp. indet. Rana catesbiana? Clemmys insculpta. Blarina sp. indet. Ænocyon dirus. Canis priscolatrans? Spilogale putorius. Brachyprotoma putorius. Boötherium bombifrons. Equus sp. indet.

Besides these forms, remains belonging to bats, various birds, snakes, and batrachians have been recognized. Of the fossils identified generically or specifically those belonging to _Megalonyx_, _Tapirus_, _Mylohyus_, _Cervalces_, _Mammut_, and _Arctotherium_ are certainly extinct. Probably, too, the bison and the species of _Felis_ are extinct. There are, therefore, pretty certainly close to 50 per cent of the species which are no longer living. This percentage and the history of some of the genera make it improbable that the assemblage belongs to the Late Wisconsin stage. Some of them could hardly have been living during the Wisconsin, when the foot of the glacier was within 100 miles toward the northeast and northwest. On the other hand, there are no species or genera present which make it necessary to refer the collection to the first interglacial. The assemblage probably belongs to the middle Pleistocene.

Coming now to the very southwestern corner of the State, we find that. _Elephas columbi_ has been met with in the bed of Hargus Creek, 3 miles above Rogersville, in Greene County (p. 150), and _E. primigenius_ on Gray’s Fork of Ten Mile Creek, near Graysville (p. 133). In the Rogersville Folio (No. 146, U. S. Geol. Surv.), Dr. F. G. Clapp described the geology of this quadrangle. On his page 10 he briefly discussed the meager Quaternary deposits of the area. These he referred to the Carmichaels formation, and indicated his opinion that it belonged to very early Pleistocene. On the geological map it is represented as occurring along Ten Mile Creek at and just below Rogersville. The occurrence of a tooth of _Elephas columbi_ just above this town and of _E. primigenius_ just above Harveys (p. 133) renders it probable that other patches of the formation exist further up the stream and along some of its branches, and that the fossils were derived from that formation. It is, of course, possible that small patches of a later deposit exist there.

Reference has been made to the Carmichaels formation. The type locality is found at Carmichaels, on Muddy Creek in Washington County. The geological description of the locality has been presented by Marius R. Campbell in the Masontown-Uniontown Folio (No. 82, U. S. Geological Survey). The formation occurs extensively along Monongahela River and other streams of western Pennsylvania. For information the reader should consult the Geological Survey Folios Nos. 144, 146, 121, 82, and 177. The deposits occur at levels considerably above the present streams and are regarded as having been laid down in old and now abandoned river channels and in tributaries of these. The time when this occurred is believed by many, if not most geologists to belong to the early Pleistocene, the Kansas stage, or possibly the Nebraskan. In the opinion of some geologists the glacial ice dammed the streams and caused their valleys to be filled with detritus. More recent Pleistocene deposits, possibly of Wisconsin age, occur at lower levels in some places south of the Wisconsin moraine; and perhaps the age of some of them has not yet been recognized. When remains of vertebrate animals are discovered, it is of great importance to determine, if possible, the exact levels of their origin.

On another page mention is made of the finding of a tooth of _Elephas primigenius_ at Lone Pine (p. 133), 7.25 miles south of southeast of Washington. This village is on Little Ten Mile Creek. No details of the discovery have been received. From Folio 144 of the U. S. Geological Survey it is learned that patches of the Carmichaels formation are found for several miles along Ten Mile Creek, near the southern boundary of the quadrangle. It seems probable that there may be patches of the same deposit along Little Ten Mile Creek, in the neighborhood of Lone Pine.

As detailed on page 70, a mastodon tooth was found many years ago about 1.5 miles south of the village of Hickory, Washington County, about twenty miles southwest of Pittsburgh. Westland Run empties into Chartiers Creek, and this into the Ohio at Pittsburgh. The geology of Burgettstown and Carnegie Quadrangles has been described by E. W. Shaw and M. J. Munn (Folio 177, U. S. Geol. Surv. 1911). No Pleistocene deposits are mapped on the stream mentioned; but just a little lower down, on Chartiers Creek, is a patch of the Carmichaels formation. Below Hickory somewhere there must be a Pleistocene deposit of some kind, and it is more probably early than late Pleistocene.

From the vicinity of Pittsburgh there have been reported remains of the mastodon (p. 69), of _Elephas columbi_ (p. 150), and of an undetermined species of elephant (p. 168). Neither of the elephants is certainly determinable. The mastodon, represented by fragments of bones and teeth, is said to have been found in the river bank, at the junction of Monongahela and Allegheny Rivers. It is impossible to determine the Pleistocene stage to which any of these proboscidean remains belong. As shown on the geological map of the Carnegie Quadrangle (Folio 177, U. S. Geol. Surv.) there are indicated here Pleistocene deposits of early, intermediate, and late stages.

Little information is furnished by a mastodon reported found on Dicks Creek in Butler County. The statements regarding the finding of elephant remains on French Creek near Meadville are vague and valueless (p. 168). Some remains of _Elephas columbi_ have been found at Tryonville, at a depth of 7 feet (p. 150). The town is on the Wisconsin moraine and the elephant probably belongs to the Late Wisconsin.

Nearly a hundred years ago a tooth of _Elephas primigenius_ was reported from a place in Erie County, called Beaverdam (p. 133). From Mr. Clyde C. Hill, civil engineer, Northeast, Erie County, the information is received that Beaverdam is a cross-roads hamlet about 23 miles south of the lake, near the prolongation of the western New York boundary line. This is within the area covered by Wisconsin drift, and it is pretty certain that the animal lived there after or near the close of the Wisconsin stage.

Just west of Erie a mastodon tooth has been found along Chase Creek (p. 70). Unless there are some unrecognized pre-Wisconsin deposits along this creek, the animal must have lived there at a time after the lake had retired to about its present limits. This would be near the very close of the Pleistocene epoch. The same conclusion must be arrived at from a study of the proboscidean remains (supposed to be those of an elephant) found at Girard.

Brief mention is made here of finds of fossil vertebrates in Pennsylvania which have not yet been mentioned; also, the localities are given where they are found, and citations of the pages where fuller descriptions are furnished:

A horse has been reported from Rutherford, Dauphin County (p. 185), and a peccary, _Platygonus vetus_ (p. 213), from Milroy, Mifflin County. Mastodons have been reported from Tunkhannock, Wyoming County; Berwick, Columbia County; Reading, Berks County; Jackson Township, York County; near Reedsville, Mifflin County; Chambersburg, Franklin County, and Bedford, Bedford County (see pp. 68, 69). _Elephas primigenius_ has been met with at Brookfield, Tioga County (p. 133); and somewhere about Chadd’s Ford, in Chester or Delaware County (p. 133).

OHIO.

(Maps 35, 36.)

The State of Ohio is partly glaciated, partly not. The unglaciated portion forms the southeastern border and constitutes close to 28 per cent of the whole surface. The glaciated area is mostly covered by the Wisconsin drift, which makes up 60 per cent of the whole surface. The remainder is covered by that part of the Illinoian drift-sheet which projects beyond the edge of the Wisconsin. This occupies about 12 per cent of the surface of the State. The unglaciated area contains Pleistocene deposits along the streams, especially along Ohio, Muskingum, Hocking, and Scioto Rivers. Probably the greater part of the materials forming these deposits were brought down the rivers which headed at the foot of the Illinoian and Wisconsin glacial ice-sheets. However, all that part of the country which was not covered by glacial ice was acted on by atmospheric agencies and suffered erosion. Hence abundant materials of non-glacial origin were swept down those tributaries of the Ohio which had their sources in the Alleghany region and down those which flowed through the unglaciated part of the State. Much of these materials was deposited along the banks of these streams and mingled with the débris from the glacial ice-sheet. Doubtless such deposits were being made during the whole Pleistocene epoch and were mostly swept away; or they may have been covered up by subsequent deposits; or the deposits of one stage may in many cases not be distinguishable from those of other stages. A perusal of chapter V of Leverett’s monograph of 1902 (Monogr. U. S. Geol. Surv., vol. XLI, 1902, pp. 228–252) and of the papers there cited, also of others published since that time, will impress the reader with the fact that an old drift, probably of Kansan or pre-Kansan age, has left traces of itself in Ohio just outside of the terminal moraine of the Wisconsin drift. This is found especially in Columbiana County; but, according to Wright (2d Geol. Surv. Pennsylvania, Z, p. 207) it extends as far westward as Canton, Stark County.

It is shown in Leverett’s paper that the streams, especially the larger ones, of southwestern New York, western Pennsylvania, and eastern Ohio had, at some time preceding that of this old drift, been deeply excavated into the underlying rocks, and that these ancient channels had become filled by the outwash from the older drift. Furthermore, terraces composed of this drift are now found along rivers of the region mentioned, at heights varying from 150 to as much as 500 feet above the present streams. Those old, deeply excavated valleys may therefore have once been filled to the highest terraces and since that time have been re-excavated to the level of the present streams. The ancient rocky floors in many cases lie now from a few to some hundreds of feet below the beds of the existing rivers. It is easily possible that the bones and teeth of early Pleistocene animals may have been buried in such valley fillings and such terrace deposits. Again, remains of such vertebrates may have been buried beneath the glacial “fringe” that has been mentioned. In such cases it may be impossible for one who is not a glaciologist, perhaps not even for him, to determine the real age of the fossils. It is, therefore, of the utmost importance that a record be kept of the exact spot where the fossil was found, so that at some future time the geology of the locality may be studied by a competent person. Naturally, other information, as that relating to the kind of deposit, depth of burial, elevation of place of burial, and the like, is valuable.

A discussion of the Illinoian drift-sheet, including that part found in Ohio, forms chapter VI of Leverett’s work of 1902 (Monogr. cit., pp. 253–291). As shown by his plate II, Illinoian drift covers a small area in the southwestern corner of the State, along Ohio River; then leaving the river and running first in a northeasterly direction, then directly north, it forms a narrow strip outside the border of the Wisconsin as far north as Richland and Holmes Counties. If it extends further east than this, it is concealed beneath the Wisconsin. It is to be expected that Illinoian drift will be discovered here and there in the greater part of the State beneath the Wisconsin where the latter shall have been penetrated in digging wells, in borings, and where streams have cut down through the later drift-sheet. In such places it will be possible to find remains of animals and plants buried in interglacial deposits laid down before the Wisconsin stage; that is, in either Sangamon or Peorian or even more remote times. On page 269 of the work just quoted, Leverett mentions a case near Lancaster, Fairfield County, where a black mucky soil was found between the Wisconsin and the Illinoian drifts. On page 273 of the same work is mentioned the occurrence of logs and pieces of wood at Bethel, Clermont County, in a gravel-bed beneath the Illinoian drift. This might be interpreted as indicating a deposit belonging to the earliest part of the Illinoian or to the Yarmouth.

The general aspects of the Illinoian drift are described by Leverett on his pages 270 to 285.

Deposits of Illinoian age may occur beyond the border of the ice-laid Illinoian drift and even beyond the Wisconsin as the result of outwash. Leverett (op. cit., p. 285) mentions the occurrence of what appears to be an Illinoian terrace along Sandy Creek, near Waynesburg, Stark County, at 70 feet above the stream, while the Wisconsin terrace is hardly 40 feet above the creek. High-level terraces are found along Licking and Muskingum Rivers from Hanover, Licking County, to McConnellsville in Morgan County, and are thought to be possibly of Illinoian age, while lower ones belong to the Wisconsin. Illinoian gravels and cobble are likewise met with along Hocking River (Leverett, op. cit., p. 288); also along the Scioto from Chillicothe nearly to its mouth. On lower-level terraces other deposits of Wisconsin age are to be looked for. Again it is seen how important it is that accurate information should be sought regarding the exact spot of interment of any vertebrate remains, as well as the elevation, the depth, and kind of materials passed through.

Map 35 has been prepared to show the distribution of the Wisconsin and Illinoian drift-sheets in Ohio. The driftless area, shown without shading of any kind, occupies the southeastern side of the State and forms a broad tract somewhat parallel with Ohio River. The Illinoian belt lies between this driftless area and the Wisconsin. Naturally it passes beneath the Wisconsin drift and probably underlies most of it. A part of the map is shaded by horizontal lines in order to show the position and extent of former Lake Maumee. This lake was an early predecessor of Lake Erie and emptied into Wabash River. The moraines laid down by the Wisconsin ice on its gradual withdrawal from the State are indicated by the stippled areas and by the letters at the sides of the map. Most of the names applied to these moraines in Ohio differ from the parts of the same moraines in Indiana. The Germantown, Eaton, and Englewood correspond to the Bloomington of Indiana; the Sidney to the Union City; the Loramie to the Salamanie; the Celina to the Wabash; and the Lima to the Fort Wayne.

Map 36 shows the localities where Pleistocene mammals have been discovered in the State and the relation of these localities to the drift-sheets and the moraines.

It is to be supposed that any animal whose remains are found in deposits overlying the Wisconsin drift lived there after the retreat of the ice-sheet from that locality. Any mastodon (maps 5, 7) that has been discovered within the area covered by the old Lake Maumee probably lived there after that lake had subsided. However, it might be possible to find along rivers, or deep cuts along railroads, animals that had lived there during Sangamon times; but this may be supposed to occur rarely. Mastodons, Nos. 34, 37, and 39 of map 7, probably lived and died after later Lake Warren had shrunken into Lake Erie.

Most of the fossil vertebrates that have been found in Ohio belong to the Late Wisconsin; that is, they lived in their respective localities after the glacial ice had retired from those localities. A few fossils may be credited to an interglacial stage, Sangamon or Peorian, which intervened between the Illinoian and the Wisconsin. Inasmuch as in the area occupied by the Illinoian drift this deposit may be cut through by rivers or railroads, it is possible that pre-Illinoian fossils might be discovered.

A tooth of _Elephas primigenius_ has been found at Waverly, Pike County, on Scioto River, as recorded on page 134. Along that river there are deposits of gravel and sand which were derived apparently from Illinoian drift, while below these Illinoian deposits is a Wisconsin terrace. The tooth above mentioned appears to have been found in a gravel-pit of the Norfolk and Western Railroad about the year 1900. The writer has not been able to secure any information as to the elevation of the pit. The elephant remains observed by Whittlesey along Scioto River, as mentioned on page 169, were probably buried in the Wisconsin terrace. A mastodon has been found in Pike County (p. 70), but the more exact locality is not recorded.

An important but apparently now lost and therefore indeterminable specimen of elephant is that to which was given the name _Elephas jacksoni_, described on page 168. It was found in the northwestern corner of Jackson County, on Little Salt Creek, probably a short time before 1838. The probability is that it was found in Wisconsin deposits, but its age is possibly greater. According to Leverett (op. cit., pp. 120, 121, 289), there are in this valley deposits which were probably laid down during the Illinoian stage. An elephant skeleton is reported to have been dug up many years ago in the village of Beverly, Washington County (p. 169), on Muskingum River. Leverett (Monogr. XLI, p. 157) states that glacial deposits belonging probably to the Wisconsin stage are found here at a height of 119 feet above the river. Inasmuch as the greater part of the village is below this level, the elephant probably belongs to Wisconsin time.

Further up the Muskingum, at or near Duncan Falls, there was found about 1857 a tooth of _Elephas primigenius_ (p. 135). The animal probably lived and died there at a time when the Wisconsin glacier was not far away. Other remains of the same species have been described from Zanesville. The bed which contained these is said to be at a height of 37 feet above the river and 20 feet from the natural surface of the ground. Inasmuch as drift outwash, believed to be of Wisconsin age, is built up here to a height of 100 feet above the river (Leverett, op. cit., p. 157), it is wholly probable that the elephant, like the one just described, lived in the vicinity of the Wisconsin ice-front. At Nashport have been discovered in swampy ground remains of _Castoroides_ (p. 273) and of _Mammut_ (p. 70). Although there is at Hanover, Licking County, across Licking River, a great dam of supposed Illinoian age and probably more or less hidden deposits of the same age along the river, the giant beaver and the mastodon just mentioned may not be older than the Wisconsin. Nevertheless, as they were found lying on gravel at a depth of 14 feet, they may have been buried there during the Sangamon stage. Along the eastern border of the State, in Columbiana County, on Salt Creek, in the southwestern part of the county, there was found, about 1845, a tooth of a horse (p. 186). It was discovered while a canal was being excavated and at a depth not to exceed 12 or 15 feet. The locality is apparently some miles south of the Wisconsin moraine. The animal lived there evidently at some time preceding the Wisconsin drift stage, possibly after the Illinoian, but quite as likely before the Illinoian. Not far away from where the horse was discovered, apparently on Little Yellow Creek, and probably not far from New Salisbury, there was found, about 1850, a fragment of the lower jaw of a tapir (p. 203). It probably lived at about the same time that the horse did. Near Millport a tooth, referred to _Elephas primigenius_, has been found (p. 135). The locality is beyond the Wisconsin moraine, but it is impossible to determine whether the beast lived there early or late in the Pleistocene.

At this point may be mentioned the discovery of remains of a peccary, supposed to be _Mylohyus nasutus_ (p. 215), and of _Mammut americanum_ (p. 70) in the southern edge of Lisbon, Columbiana County, apparently along Middle Fork of Little Beaver River. This locality is on the border of the Wisconsin drift-sheet, and the peccary and the mastodon might well have lived there with the horse and the tapir mentioned above.

Not many localities within the area of the Illinoian drift in Ohio have furnished vertebrate fossils.

Lyell in 1843, as stated on page 71, reported that teeth of mastodons and of elephants had been found on the Cincinnati side of the river, on the high terraces.

From Professor N. M. Fenneman the writer learns that Lyell’s reference could hardly apply to any other locality than Terrace Park or Milford. Here are found some fragments of an Illinoian terrace that would hardly be spoken of casually as such, while the Wisconsin deposit is present as an upper and a lower terrace.

In Hyde Park, as detailed on page 71, considerable parts of a mastodon and some remains of a horse (p. 185), probably _Equus complicatus_, have been discovered. The age of these remains certainly antedates that of the Wisconsin; and it is not improbable that the excavation was carried through the Illinoian drift into an older and probably interglacial deposit. Professor Fenneman writes that this area is only thinly covered by Illinoian drift and is also far beyond the limits of the Wisconsin outwash.

The occurrence of _Bison latifrons_ near Fincastle, in Brown County (p. 257), must be noted. The fine pair of horn-cores now in the Cincinnati Society of Natural History may have been buried in deposits of Sangamon age. It is not, however, impossible that they were in an interglacial bed below the Illinoian drift.

On page 135 there has been given an account of the finding of a skull of _Elephas primigenius_, somewhat more than a mile east of New Burlington. The locality is treated in proper detail in N. M. Fenneman’s paper entitled “Geology of Cincinnati and Vicinity” (Bull. 19, Geol. Surv. Ohio, p. 158). According to this account the skull was buried in a lacustrine silt laid down probably when the Wisconsin glacier was not far away from that region. The surrounding country is covered with Illinoian drift. This skull is now the property of the U. S. National Museum.

In the collection of the Ohio State Archæological and Historical Society at Columbus there are remains of _Platygonus compressus_, jaws and good teeth, which were found about a mile north of Chalfants, in Perry County, and along Jonathan Creek. This place is within the area covered by Illinoian drift. It is possible that the remains are as old as the Sangamon, but it is also possible that they belong to the close of the Wisconsin stage (p. 215).

The writer knows of no other fossil vertebrates that have certainly been found within the area occupied by the Illinoian till as a surface deposit.

As shown by map 36, by far the larger number of Pleistocene vertebrates which have been discovered in Ohio have been met with within the region occupied by the Wisconsin drift-sheet. One reason for this preponderance is the greater area included. Another reason may be found in the fact that the conditions were more favorable for the preservation of teeth and bones. Much of the country was flat and swampy and the bones buried in clay and muck have always been soaked with water. Also there has been less erosion going on. Erosion leads to exposure and therefore to destruction of skeletons.

On the map referred to are shown the various moraines that were left by the Wisconsin ice-sheet in its retreat toward the north. Inasmuch as most of the burials were in swamps resting on the drift, the animals must have lived and died there after the ice had left that vicinity; how long after one may not be able to determine. The mastodons and elephants which have been found close to the shore of Lake Erie, especially if buried near the surface, must have lived there at or after the time when the waters had shrunken into Lake Warren. Such cases are furnished by the mastodons and elephants found at Amboy (east of Ashtabula) (pp. 137, 150), at Cleveland (p. 79), and in Brownhelm Township, in Lorain County (p. 79). The town of Amboy is about 130 feet above lake level and the gravel-pit which there furnished _Elephas primigenius_ and _E. columbi_ was probably at about the same level. The writer has not been able to confirm any case in which remains of proboscideans have been met with on the south shore of the lake at a level lower than the Warren beach. Mastodons may be traced to a lower level at the western end of the lake. The one found in Springfield Township, Lucas County (p. 77), was buried in deposits only about 45 feet above Lake Erie. As shown by the topographical maps, the descent from this place and from Bowling Green, Wood County, to the lake is a gradual one. It may become possible to follow the presence of the mastodons, the elephants, and the giant beaver in Ohio up to the time when the lake assumed its present level.

For information regarding the several interesting discoveries of the giant beaver (_Castoroides ohioensis_) pages 273 to 275 may be consulted.

It is hardly necessary to take up one by one all the cases of vertebrates that have been met with within the area covered by Wisconsin drift. With the few exceptions noted below, their geological age is usually to be regarded as Late Wisconsin. Along the southern border of this drift, where the remains are deeply buried, it is not unlikely that they lie in a pre-Wisconsin interglacial deposit. Along Great Miami and Muskingum Rivers there is always a possibility that the fossils may occur in a terrace or in a deep valley deposit of Illinoian age.

About a mile east of Overpeck, Butler County, there has been found the skull of an extinct bear, _Ursus procerus_ Miller (Hay, Geol. Surv. Indiana, vol. XXXVI, 1912, pp. 772–776, figs. 71–73). It was found at a depth of 28 feet and about 3 or 4 feet above the limestone rock of that region. To the writer it seems quite certain that the Wisconsin drift had been penetrated and that the skull was in either a Sangamon interglacial deposit or something still older.

Columbus furnishes one of the rare cases in which horse remains have been found within the Wisconsin glaciated area (p. 186). We are then required to determine whether or not the horse, _Equus complicatus_, did not live there after the close of the Wisconsin stage. As said on the page cited, the first remains of horses discovered at Columbus were reported as having been found in crevices of the limestone and in the red clay filling such fissures. An examination of the Columbus Folio (197, U. S. Geol. Surv., p. 8) will show that in such crevices, south of Scioto River, a red clay is found which antedates the Illinoian drift, so that one might fairly refer the horse remains reported by Whittlesey to a pre-Illinoian interglacial stage, possibly the Aftonian. The horse-teeth found in the excavations at the penitentiary close to Scioto River may be as old as those found in the rock fissures, or they may have been buried in a post-Illinoian interglacial deposit. Such deposits have been found at various places in the quadrangle (fol. cit., p. 9).

As to the peccaries discovered at Columbus (p. 214), the writer sees no reason why they should not be regarded as belonging to the Late Wisconsin.

MICHIGAN.

To understand the Pleistocene geology of the southern peninsula of Michigan, it is indispensable to study Monograph LIII of the U. S. Geological Survey, by Frank Leverett and F. B. Taylor. The whole peninsula is overlain by glacial deposits laid down by the Wisconsin ice-sheet. A glance at their glacial map (plate VII) will indicate to the student the complexity of glacial problems in this region. The ice invaded the State from three sides: on the west from Lake Michigan, on the east from Lake Huron, and on the southeast from Lake Erie.

On the west, close to Lake Michigan, is a system of Lake-border moraines. This system has been traced more or less satisfactorily around to Saginaw Bay, Lake Huron. A little farther out, in the southwestern corner of the State, is Valparaiso moraine. This extends nearly to the northern end of the peninsula, where it connects with Charlotte moraine system. Farther in than the Valparaiso system is the Kalamazoo. This extends northeastwardly from the Indiana line to Barry County, where it turns east and at Jackson joins the Mississinawa system reaching northeast from the northeastern corner of Indiana. The Valparaiso and Kalamazoo moraines are in places closely associated. The attack on the eastern side of the State came principally from a lobe which flowed through Saginaw Bay. Nearest Lake Huron, following it around from Port Huron to the northern end of the peninsula and then turning west, the Port Huron moraine connected with the moraine along Lake Michigan.

Farther inland is the Charlotte system. On the north, just above latitude 44°, this joins the Valparaiso moraine, runs southward west of Lansing, then turns eastward, then northeastward, and connects with the Defiance moraine, which passes around the western end of Lake Erie. Reaching far out from the head of Saginaw Bay, and concentric with it, to Hastings, 100 miles away, are many minor moraines.

Besides the Wisconsin drift which forms the surface deposit in Michigan, there are, according to present indications, one or more pre-Wisconsin drifts. Leverett (Monogr. LIII, p. 72) mentions several localities where what appears to be more indurated till is encountered, sometimes at a depth of 100 feet. Taylor (op. cit., pp. 289–290) states that “a till older than that deposited by the Wisconsin ice-sheet seems to underlie more or less continuously all of the later, or Wisconsin, drift in Indiana and the southern peninsula of Michigan.” Along the western shore of Lake Huron, north of Port Huron and along the streams, as reported by Taylor (p. 290), there are several exposures of Illinoian till, in some cases as much as 30 to 50 feet thick. In one case there is an old soil at the top of this till. In such old soils it may be possible to find fossil vertebrates of Sangamon or Peorian times, horses for example.

The fossil vertebrates found up to the present time in Michigan are not numerous in species or individuals; all appear to belong to the middle or late Wisconsin times. A peccary, _Platygonus compressus_, has been found at Belding, Ionia County (p. 215). Two musk-oxen have been discovered in the State. At Manchester, Washtenaw County, has been found a fine skull of _Symbos cavifrons_ (p. 250). At Moorland, Muskegon County, was obtained a skull which has been called _Boötherium sargenti_.

Details regarding the mastodons which have been found in Michigan are given on pages 80 to 88. Only two localities in the State have furnished remains of _Elephas primigenius_. These are Three Oaks, Berrien County (p. 137), and Eaton Rapids, Eaton County (p. 137). _Elephas columbi_ has been encountered only once in the State, as far as is known; this was in the northern part of Jackson County (p. 151).

Elephants belonging quite certainly to either _E. primigenius_ or _E. columbi_, but for one reason or another not determined, have been found in four localities. These are East Saginaw, Saginaw County; Macomb County; Grand Ledge, Eaton County; and Buchanan, Berrien County. (See page 171.)

The giant beaver, _Castoroides ohioensis_, found a congenial home in the swamps of southern Michigan in the late Pleistocene. It has been met with somewhere in Berrien County; at Adrian, Lenawee County; at Ann Arbor, Washtenaw County; at Attica, Lapeer County; and at Owosso, Lapeer County (pp. 275–276).

INDIANA.

(Map 37.)

Whoever wishes to gain a knowledge of the Pleistocene geology of Indiana, as it is understood to-day, must study Leverett’s two great treatises, forming Monographs XXXVII and LIII of the U. S. Geological Survey. The first is entitled “The Illinois Glacial Lobe,” and was published in 1899; the second has the title “The Pleistocene of Indiana and Michigan and the History of the Great Lakes.” The portion of the latter monograph which deals with Michigan was written by F. B. Taylor. On pages 33 to 54 is a very full bibliography of the subject, consisting of about 400 titles.

From the glacial map of Monograph XXXVIII, plates V and VI, the writer has prepared map 37. This shows which part of the State has escaped glaciation, which has been subjected to the action of the Illinoian ice-sheet, and which has been covered by the last, or Wisconsin, glacial ice. It will be seen that about one-sixth of the State, that forming an irregular triangle whose apex is in Brown County and whose base is formed by the Ohio River, has never been covered by glacial ice. North of this is a bilobed area which is covered by till of Illinoian age. The rest of the State (somewhat less than two-thirds of it) is overlain by the débris left by the Wisconsin ice-sheet and subsequent deposits.

This northern area is to a great extent occupied by belts called moraines, along which the materials are usually coarse, often full of boulders, and frequently standing at a higher level than the surface on each side of them. These moraines show where for long periods during its retreat, or perhaps sometimes its advances, the ice-sheet paused and piled up a part of its load of rocks, gravel, and sand. It will be noticed that these moraines are somewhat concentric. On the right of the map are seen those moraines which were left by the ice-lobe which came down Lake Erie and later retired in that direction. Around the southern end of Lake Michigan are the moraines laid down by the ice of the Michigan lobe. The latter will be better seen on a glacial map of Illinois (Monogr. U. S. Geol. Surv. XXXVIII, plate VI). In their advance the two lobes met and coalesced and produced more or less irregular and anastomosing moraines.

On the right hand the moraines of the Erie lobe pass on into Ohio, where, however, they have often been given other names. On the left the moraines of the Lake Michigan lobe continue into Illinois and retain the same names. Both groups of moraines are prolonged into the southern peninsula of Michigan.

On account of the comparatively recent recession of the Wisconsin ice-sheet, the surface has not become eroded sufficiently to drain away the water which was left in depressions of the surface. A large part of Indiana is, or has been until recently, covered by swamps, lakes, and ponds, and in such localities the bones and teeth of vertebrate animals are best preserved during the early stages of fossilization. For this reason great numbers of teeth and bones, sometimes nearly whole skeletons, are met with in draining these swamps.

The southern border of the Illinoian drift, beginning at Cincinnati, follows Ohio River on the Kentucky side to Jeffersonville, then passes west of north into Brown County, whence, turning southwest, it strikes the East Fork of White River in Du Bois County; thence, following White River a short distance, it crosses the Wabash in Posey County. Northward, along this terminal moraine (map 37, figs. 1, 2) of the Wisconsin drift, the Illinoian, passing beneath this, disappears from the surface.

The surface of the Illinoian area is better drained than the Wisconsin area. Fewer fossils are found, and on various accounts they are of less value. Usually the exact locality and kind of deposit is not recorded. They may be found washed out of river and creek banks and may have in reality been buried in sediments that were laid down in Wisconsin times by the streams that carried away the mud, sand, and gravel from the glacier. The driftless area has been exposed for many geological ages to the influence of physical and chemical agencies. Its surface is, therefore, more diversified by hills and valleys and streams. In the limestones of this region caves are likely to be found, and these now and then furnish fossil bones and teeth.

During more than one of the glacial stages, perhaps during the earliest, the Ohio has served as the drainage-way for the waters that escaped from the glacial front. This subject is discussed by Leverett in Monograph XLI of the U. S. Geological Survey. As a result of this conveyance of glacial waters, the great trough of this stream may contain here and there deposits of the Illinoian stage or even of older deposits. Remains of _Megalonyx_ (p. 32) and of a horse (p. 186) have been found in the right bank of the Ohio, at Evansville, Indiana. At Bigbone Lick, Kentucky, close to the Ohio, horses have been discovered, _Mylodon_ and _Megalonyx_. These seem to occur in Sangamon interglacial beds overlying the Illinoian.

The Illinoian drift, probably everywhere in central and northern Indiana, underlies the Wisconsin. For some miles back from its terminal moraine the Wisconsin drift is thin; and possibly the Illinoian may. be found exposed in creek or river banks, or in railroad cuts. Furthermore, Leverett (Monogr. LIII, p. 72) writes:

“Probably a considerable number of the heavy deposits of drift in central and northern Indiana are of pre-Wisconsin age, but as they are largely sand and gravel, or loose-textured material, they can not easily be discriminated from the Wisconsin.”

Such deposits are likely to be covered by only a thin layer of Wisconsin till. In many places in Indiana there have been found, deep down in the drift, old soils, muck beds, and vegetation in various forms. These beds appear to indicate interglacial deposits, most probably the Sangamon. Now, various genera of vertebrates, among them horses, tapirs, and mylodons, are not known to have existed after the Wisconsin glacial stage. If, however, remains of such animals should be collected in central or northern Indiana, or Ohio, or in southern Michigan, they might be reported as having been found in late Wisconsin beds, when really they had been derived from pre-Wisconsin interglacial soils.

It is interesting to observe that when the Wisconsin ice-sheet began to withdraw lakes began to form along its borders. One of these, Lake Chicago, appeared at the south end of the present Lake Michigan and for a long time discharged its waters down Illinois River. Another, Lake Maumee, occupied the basin of Maumee River as far west as Fort Wayne, and emptied down the Wabash. For details connected with the close of the Pleistocene in the region of Lake Michigan the reader should consult Frank C. Baker’s “The Life of the Pleistocene, or Glacial Period” (Univ. Ill. Bull. XVII, 1920).

A brief mention will be made here of the principal Pleistocene vertebrates that have been found in Indiana; also the localities where found, together with citations of the pages where fuller information is furnished.

The ground-sloth _Megalonyx_ has been collected near Evansville (p. 32). With it were secured remains of an undetermined bison (p. 257), a Virginia deer (p. 228), a horse (p. 186), a tapir (p. 203), and the dog _Ænocyon dirus_ (p. 32). Peccaries have been found in Gibson County (p. 216), in Wabash County (p. 218), and two species at Williams, Lawrence County (p. 217). At the same place was discovered the shell of a box-tortoise. Remains of deer have been discovered somewhere in Vanderburg County, including the existing deer and an extinct species, _Odocoileus dolichopsis_; at Harrisville, Randolph County; and at Roann, Wabash County. Bisons of an extinct species have been secured at Vincennes (p. 258).

The existing bison appears to have been found in Jasper County (p. 268). Of musk-oxen, _Symbos cavifrons_ has been collected at Hebron, Porter County (p. 252); at Wailesboro, Bartholomew County (p. 251); somewhere in Randolph County (p. 252); and probably in Beaver Lake, Newton County (p. 252). The existing musk-ox, _Ovibos moschatus_, has been discovered near Richmond (p. 252).

Mastodon remains are not uncommon, especially in the northern half of the State. It is hardly to be supposed that these animals were more abundant there during the late Pleistocene than in many other places in the region east of the Mississippi. The conditions for their preservation were evidently more favorable there than anywhere else, unless in Orange County, New York. Burial in swamp mud kept the bones from decay; and the imperfect drainage protected them from destruction by erosion. The various finds are described on pages 88 to 100.

Elephants are less well represented in Indiana than are the mastodons, but are not rare (pp. 138, 151, 171). Two species were present in the State, _Elephas primigenius_ and _E. columbi_. Beavers were doubtless abundant, but there appears to be no definite record of any find. However, the giant beaver has been recorded from several localities (pp. 276 to 278).

The great extinct dog _Ænocyon dirus_ was first found near Evansville (p. 32), and the coyote, _Canis latrans_, has been reported from Boone County. The latter is said to have been found in association with the mammoth (Cope and Wortman, 14th Ann. Rep. Geol. Surv. Indiana, p. 7).

ILLINOIS.

(Map 38.)

As a foundation for a knowledge of the Pleistocene geology of Illinois, the student must take Leverett’s work entitled “The Illinois Glacial Lobe.” This is Monograph XLVIII of the U. S. Geological Survey, a volume of 817 pages, with maps and figures. For a knowledge of the changes which occurred around the south end of Lake Michigan on the retirement of the Wisconsin glacier, see Dr. Frank C. Baker’s work, “The Life of the Pleistocene, or Glacial, Period” (Univ. Ills. Bull. XVII, 1920).

Illinois is eminently a glaciated State, as is to be recognized on Leverett’s plate VI. A little triangle in the northwestern corner, comprising about 600 square miles, and an irregular tract of perhaps 3,000 square miles at the southern end of the State constitute the whole of the unglaciated area out of 56,650 square miles. Two glacial stages are prominent, the Wisconsin and the Illinoian. The first was laid down by the Lake Michigan lobe, which sent its icy mass southwestward as far as Shelbyville. Westward the border moraine extends to Peoria, then north to west of Princeton, then northeast to enter Wisconsin 55 miles west of Lake Michigan. Eastward, of course, the deposits of till and the moraines extend into Indiana. North of the Shelbyville moraine is the Champaign. A more powerful moraine is the Bloomington, which forms a loop through the State, extending from Danville, Illinois, through Bloomington to Peoria, where it appears to have overridden the Shelbyville and thence northward, forming the outer border of the Wisconsin drift area. North of this moraine is located that called the Marseilles, while sweeping around the south end of Lake Michigan into Indiana and Michigan is the Valparaiso system.

South and west of the area of the Wisconsin drift is the Illinoian. At Mount Vernon the border crosses the Wabash and traverses Illinois, striking the Mississippi River at Carbondale. It then follows the Mississippi north to a point above Keokuk, where it enters Iowa. It reenters Illinois between Rock Island and Clinton and extends into Wisconsin.

On Leverett’s map (Monogr. XXXVIII, plate VI) there is indicated in northern Illinois, between the Illinoian and the Wisconsin, a tract supposed to belong to the Iowan; but Alden (U. S. Geol. Surv., Prof. Pap. 106, 1918, p. 173) holds that there is no good evidence that the Iowan extends into southern Wisconsin and Illinois. The supposed Iowan (op. cit., plate III) is mapped as Illinoian.

The glacial stage which preceded the Illinoian is the Kansan. This in Iowa extends eastward to the Mississippi, and one might naturally expect that it would be found underlying the Illinoian east of the river. Leverett (Monogr. XXXVIII, p. 105) presents evidences of its presence in western Illinois. Among these evidences is the presence in Hancock and Adams Counties of another till sheet below the Illinoian and separated from it by a black soil. This Kansan or some other pre-Illinoian till sheet has been found in many places in Illinois (op. cit., pp. 107–118).

Animal remains are not likely to be inclosed in the materials of the moraines or of the intermorainal till; but this is possible. A musk-ox or a hairy mammoth might have died not far away from the foot of a stationary or advancing glacier and its bones might have become incorporated in the moraine. Furthermore, inasmuch as any glacial stage began while the glacier was yet in the far north and ended only when it got back there, many non-glacial deposits belonging to that glacial stage were probably laid down south of it; and it would be difficult or impossible to distinguish these from interglacial deposits. However, it was these deposits which were laid down after the glacial ice had withdrawn, whether glacial or interglacial, which are of more interest to the palæontologist, because in them are to be found the fossil remains of animals and plants.

The last of the interglacial stages, that which immediately preceded the Wisconsin and followed the Iowan, is known as the Peorian. This takes its name from a locality a few miles east of Peoria (Leverett, Monogr. XXXVIII, p. 187). Here the Shelbyville till sheet is underlain by a bed of fossiliferous loess from 8 to 12 feet in thickness. Beneath the loess is fully 100 feet of Illinoian drift. This loess seemed to the geologists who examined it to be a deposit of more recent date than the Sangamon.

The Peorian interglacial stage and the preceding Iowan glacial stage have received much attention within recent years. In 1917 (Geol. Surv. Iowa, vol. XXVI, pp. 49–212), Alden and Leighton presented the results of their studies on the Iowan drift and the loess associated with it. In 1918 (U. S. Geol. Surv., Prof. Pap. 106, pp. 1–356), Alden dealt with the Quaternary geology of southeastern Wisconsin. The results of these investigations have been to establish the fact that a sheet of till intermediate between the Illinoian and the Wisconsin had been laid down, that which had already been designated as the Iowan; furthermore, that immediately following this there was deposited a covering of loess. It was further concluded that this is the main loess deposit, much of what has been regarded as Sangamon loess being really loess of a later stage, the Peorian.

As no Iowan drift is known to be present in Illinois to separate the loess of the Sangamon from the Peorian, it must be difficult, often impossible in our present state of knowledge, to distinguish the one from the other. The Sangamon loess was laid down probably long after the Illinoian ice disappeared, so that there was time for the Illinoian drift to become leached and otherwise modified and for the accumulation of old soils and peat-beds.

On the other hand, the old soils of the Peorian stage are likely to overlie the loess. Unfortunately, the desired indications of geological age are not always present where bones and teeth are found; or, if present, are not always observed. We must, therefore, make our assignments of fossils to one stage or the other with great circumspection or leave the decision in abeyance.

Reference has already been made to the presence of Kansan drift in western Illinois and of black soils intervening between it and the Illinoian. Such soils must be referred to the Yarmouth interglacial stage. Whether or not still older glacial or interglacial deposits occur in Illinois is problematic.

In Illinois any considerable number of species of fossil vertebrates are rarely found together. The localities are widely scattered and a single species or two in each is the rule (map 38). In later glacial deposits around the south end of Lake Michigan have been discovered the dogfish _Amiatus calvus_ and a sun-fish belonging to the genus _Lepomis_. Baker (Univ. Ill. Bull. XVIII, p. 85) reported the humerus of the merganser, _Mergus serrator_, from the same region. The ground-sloth _Megalonyx jeffersonii_ (pp. 33–34) has been found at Urbana, Galena, and Alton.

The few horses are described on page 187. Peccaries have been found at three localities (p. 218). For the specimens of deer that have come to light, see page 229. A species of _Cervalces_ and the moose _Alces americanus_ have been met with in Will County (p. 107). The reindeer has been recognized from poor materials found at Alton. The prong-horn _Antilocapra_ appears to have lived in the region of Galena, as shown by Wisconsin specimens. The remarkable antelope _Taurotragus americanus_ has been found at Alton (p. 339). As to the musk-oxen and the bisons, the reader may refer to pages 251, 259, 268; for the mastodons and elephants, to pages 100, 140, 152, and 176.

Of the rodents, the muskrat has been found about Chicago; the pocket gopher at Alton and Galena; the ground hog at the same places (p. 343). The beaver (p. 339) likewise occurs at Alton. The giant beaver, _Castoroides ohioensis_, has been collected at four widely removed places (p. 279). The rabbit, _Sylvilagus floridanus_, was included among the animals found in the lead crevices of the region about Galena, where also have been found an extinct species of raccoon, _Procyon priscus_, what appears to be a large dog _Canis_ (or _Ænocyon_) _mississippiensis_, the coyote, _Canis latrans_, and the fox _Urocyon cinereoargenteus_. The bear, _Ursus americanus_, and the common gray wolf, _Canis nubilus_, appear to have existed in the middle Pleistocene at Alton.

A skull of _Felis couguar_, the yet existing panther or mountain lion, has been found in Randolph County, in the bed of Kaskaskia River. It probably belongs to the late Pleistocene.

A considerable fauna has been secured in the lead region about Galena, in Illinois, Iowa, and Wisconsin. The collectors and describers of this were not careful to designate the localities, and in some cases these can not at present be determined. These collections are discussed on page 343, in the account of the geology of Wisconsin.

An interesting list of Late Wisconsin mammals has been secured near Whitewillow, Kendall County. From Dr. E. S. Riggs, of Field Museum of Natural History, and from Netta C. Anderson’s list, the writer learns that at least six skulls of the common mastodon, together with many other parts of the skeleton, has been taken from a well 10 feet in diameter (p. 109). Above, there were bones of bison (p. 269), deer (p. 229), and elk (p. 240). It is stated that a layer of these about 2 feet thick was encountered at a depth of about 5 feet.

Mr. George Langford, of Joliet, states that he made a collection of bones 15 miles west of Joliet and 5 miles west by north of Minooka. The more exact locality he gave as township 35 north, range 8 east, and probably section 27, on the farm of John Bamford. Apparently both Riggs and Langford obtained their materials at the same spot. The latter has sent the writer some bones from this place, including those of _Cervalces_, _Alces americanus_, and a leg-bone of some undescribed species of sheep or goat. He also reported the finding of the elk. For other remarks see page 269. This locality is in the region mapped by Leverett as having been occupied, after the retirement of the Wisconsin glacial ice, by temporary lakes. The presence of the moose here seems to indicate a climate somewhat severer than that now prevailing in that region. Since the occupancy of the country by the European race the moose has not been known to come further south than northern Wisconsin. The list of species obtained is as follows: _Mammut americanum_, _Ovis_ sp. indet., _Odocoileus virginianus_, _Cervus canadensis_, _Alces americanus_, _Cervalces roosevelti?_.

A brief description of the bone referred to _Ovis_ is presented. The lower epiphysis is missing, but an allowance is made for this (fig. 13).

_Comparisons of the metatarsals of a sheep, of a goat, of Næmorhedus, and of Orvis sp. from Whitewillow, in millimeters, together with indices in one-hundredths of the length._

┌──────────────┬───────┬─────╥───────┬─────╥──────┬─────╥───────┬─────┐ │ Measurements │ │ ║ Capra │ ║ │ ║White- │ │ │ taken. │ Næmo- │Indi-║hircus │Indi-║ │Indi-║willow │Indi-│ │ │rhedus.│ces. ║155623.│ces. ║Sheep.│ces. ║animal.│ces. │ ├──────────────┼───────┼─────╫───────┼─────╫──────┼─────╫───────┼─────┤ │Length on │ │ ║ │ ║ │ ║ │ │ │ outer border│ │ ║ │ ║ │ ║ │ │ │ of bone │ 170│ 100║ 120│ 100║ 152│ 100║ 185±│ 100│ │Side-to-side │ │ ║ │ ║ │ ║ │ │ │ width of │ │ ║ │ ║ │ ║ │ │ │ upper │ │ ║ │ ║ │ ║ │ │ │ articular │ │ ║ │ ║ │ ║ │ │ │ surface │ 36│ 21.2║ 23│ 19.2║ 23│ 15.1║ 37.5│ 20.3│ │Fore-and-aft │ │ ║ │ ║ │ ║ │ │ │ width of │ │ ║ │ ║ │ ║ │ │ │ upper │ │ ║ │ ║ │ ║ │ │ │ articular │ │ ║ │ ║ │ ║ │ │ │ surface │ 30.5│ 17.4║ 20│ 16.7║ 21│ 13.8║ 37.5│ 20.3│ │Side-to-side │ │ ║ │ ║ │ ║ │ │ │ width, at │ │ ║ │ ║ │ ║ │ │ │ middle of │ │ ║ │ ║ │ ║ │ │ │ length │ 23│ 13.5║ 15│ 12.5║ 14│ 9.2║ 19.0│ 10.3│ │Fore-and-aft │ │ ║ │ ║ │ ║ │ │ │ width at │ │ ║ │ ║ │ ║ │ │ │ middle of │ │ ║ │ ║ │ ║ │ │ │ length │ 17.5│ 10.3║ 11.5│ 9.1║ 13│ 8.6║ 20.0│ 10.8│ │Side-to-side │ │ ║ │ ║ │ ║ │ │ │ width at │ │ ║ │ ║ │ ║ │ │ │ lower end │ │ ║ │ ║ │ ║ │ │ │ just above │ │ ║ │ ║ │ ║ │ │ │ epiphysis │ 38│ 22.4║ 27│ 22.5║ 27│ 17.8║ 35.0│ 19.5│ │Side-to-side │ │ ║ │ ║ │ ║ │ │ │ width across│ │ ║ │ ║ │ ║ │ │ │ lower │ │ ║ │ ║ │ ║ │ │ │ articular │ │ ║ │ ║ │ ║ │ │ │ surface │ 41│ 24.1║ 27.2│ 22.5║ 25│ 16.4║ │ │ └──────────────┴───────┴─────╨───────┴─────╨──────┴─────╨───────┴─────┘

From Alton, the U. S. National Museum has come into possession of a collection which furnishes 15 species of fossil mammals. This was made some time before 1883 by Hon. William McAdams, of Alton. It was briefly mentioned by him at the Minneapolis meeting of the American Association for the Advancement of Science in 1883 (Proceedings, vol. XXII, p. 268). Apparently the collection was secured for the U. S. Geological Survey by Professor O. C. Marsh and remained at Yale University until after his death. The species were described by the writer in 1920 (Proc. U. S. Nat. Mus., vol. LVIII, pp. 109–117). This collection seemed especially valuable because the species were found inclosed in supposed nodules of loess. In our country the loess has furnished few such remains. The following is the list of the species as determined. Those marked by a dagger are extinct.

†Megalonyx jeffersonii (p. 33). †Equus sp. indet. (p. 187). †Platygonus cumberlandensis? (p. 219). †Sangamona fugitiva. †Cervalces roosevelti?. †Rangifer muscatinensis? (p. 246). †Taurotragus americanus. †Symbos cavifrons (p. 254). †Bison sp. indet. (p. 259). †Mammut americanum (p. 102). Castor canadensis. Marmota monax. †Castoroides ohioensis (p. 279). Geomys bursarius. Ursus americanus.

Of these 15 species at least two-thirds are now extinct. This large number might appear to indicate that the time of their existence was rather early in the Pleistocene. However, it is quite certain that the loess belongs somewhere about the middle of the Pleistocene; and there are no species that require an earlier date.

After the writer’s descriptions of the fossils had been published, an important paper on the geology of the locality was issued (Jour. Geol., vol. XXIX, 1921, pp. 505–514) by Professor Morris M. Leighton, who had been commissioned by the Illinois Geological Survey to visit and study the deposits involved. With the aid of Mr. John D. Adams, son of the collector of the mammalian fossils, Professor Leighton succeeded in finding the quarry in which most of the fossils had been collected.

At one quarry in Alton Professor Leighton obtained the following geological section, the description of which is here somewhat abridged:

_Feet._ Soil loessial, dark brown, leached 1 Loess, brown above, grading below into buff, leached 4 to 5 feet, maximum thickness 20 Loess distinctly more reddish than that above; many fossil snails, thickness about 30 Glacial till, reddish, with pebbles of Canadian rocks; more oxidized than overlying loess; thickness 1–3 Mississippian limestone, about 100

The concretions which hold the mammalian fossils were found to lie between the upper surface of the till and the overlying loess; occasionally a concretion bears a drift pebble. The concretions have resulted from the lime which in solution was brought down from the loess and again precipitated so as to cement the loess materials around the fossils.

Professor Leighton was not able to determine definitely the ages of the till and of the two deposits of loess. As to the till, its geographic location suggested that it belonged to the Illinoian, but it had many of the characteristics of the Kansan. The latter is believed to be present at St. Louis and other localities not far away. Before the overlying reddish loess had been deposited the till had suffered weathering and erosion, indicating a considerable lapse of time had intervened. The lower reddish loess presented many evidences that it is a deposit distinct from the upper buff loess; and there seemed to be some indications of at least a short interval between them. Leighton’s conclusion was as follows:

If the drift is Kansan in age, the reddish loess may be Sangamon; if, on the other hand, the drift be Illinoian, the reddish loess probably is Peorian. It is unlike any Peorian loess of which the writer knows, but the color does not necessarily preclude that possibility.

As to the upper loess, Leighton thought it might be of early Peorian age, but possibly of early Wisconsin. However, his final conclusion was thus expressed:

“If the till proves to be Kansan in age, the weathering of the drift may be credited to the Yarmouth interglacial epoch, the mammalian fauna to late Illinoian or early Sangamon times, the reddish loess probably to the Sangamon, and the buff loess to the Iowan.... However this may be, the Illinoian and Sangamon epochs are post-mid-Pleistocene from the standpoint of duration of the Pleistocene and the fauna represented by the McAdams collection may be regarded as post-mid-Pleistocene.”

WISCONSIN.

The greater part of this State is covered by the drift-sheet which has derived its name from the State, but in the southwestern corner is a considerable tract which has never been subjected to glacial action. A small part of this area extends southward into Illinois and another part into northeastern Iowa. In Wisconsin it reaches eastward to Baraboo. East of this driftless area is a tract lying along the southern border of the State and reaching eastward about to 88° 40′ longitude, which is covered by the Illinoian drift.

The most detailed geological survey of any part of Wisconsin, so far as regards the Pleistocene, is that made by Dr. W. C. Alden, of the U. S. Geological Survey, of the area comprised between the boundary of the State on the south and 44 degrees of latitude on the north and between Lake Michigan on the east and 90 degrees of longitude on the west. On the western side it joins the Mineral Point Quadrangle, to be mentioned further along. There is, therefore, a wide strip surveyed across the whole State. The area treated by Alden is, of course, nearly entirely covered by Wisconsin drift. In the southwestern corner a considerable part of the driftless region is included. East of this, as already stated, is a tract which the Wisconsin ice-sheet did not reach and which shows Illinoian ground moraine and some terminal moraines of Illinoian drift. This narrows as it approaches its eastward limit.

Alden (p. 166) informs us that at no place in the area subjected to vigorous glaciation by the Wisconsin ice-sheet had soils or vegetal deposits been found between the Wisconsin drift and the earlier drifts. At several places, however, deposits have been discovered which probably belong to earlier glacial stages. Just outside the area mapped by Alden, in Calumet and Outagamie Counties, Lawson (Bull. Wis. Nat. Hist. Soc., vol. II, pp. 170–173) has recorded the discovery of much wood and other vegetable matter. Baker (“Life of the Pleistocene,” p. 317) has referred the deposits to the Sangamon. These interglacial deposits of uncertain age need not be here noted further. In this Wisconsin area some remains of mastodons and elephants have been met with, but all are relics of a time after the partial or complete recession of the Wisconsin glacier. Remains of two individuals of _Elephas primigenius_ have been found in Milwaukee (p. 143). It is evident that they lived there after the withdrawal of the Wisconsin ice-sheet. One of these was buried beneath peat and clay at a depth of 10 feet or more and at a level of about 100 feet above the present level of Lake Michigan.

At Dover, in Racine County, in 1878, a proboscidean tusk and some bones were found in a peat-bog. They have been identified as those of a mastodon, but of this one can not be certain. The age of the deposits is that of the Late Wisconsin stage, after the withdrawal from that vicinity of the ice, but how long after one can not say. The Milwaukee Public Museum has a tooth of a mastodon (p. 111), labeled as found at Waukesha. Its geological age is that of the other remains here referred to. In the collection of the University of Wisconsin is a large vertebra of a proboscidean which was found in Lake Monona. Its time of burial must have been late Wisconsin. Inasmuch as no remains of vertebrate animals have yet been found in Wisconsin, in the area covered by the Illinoian drift, it is not necessary to dwell on this region. It is not certain that there is beneath it a still older drift; but there are, according to Alden, some indications of such deposits.

For a knowledge of the driftless area, first of all, may be consulted the report made by Chamberlin and Salisbury in 1885 (6th Ann. Rep. U. S. Geol. Surv., pp. 199–322, with plates). Alden’s work above referred to maps a part of the region. Grant and Burchard have studied the geology of the Lancaster and Mineral Point Quadrangles (Folio U. S. Geol. Surv. 145). Their text-figure 1 is here reproduced, inasmuch as it shows the relation of the region to the surrounding glaciated areas (fig. 14). The topographical map of Folio 145 and that of Chamberlin and Salisbury will show the uneven character of the surface. This has resulted from the erosion undergone during the whole of the Pleistocene. Much of the area is covered with a coating of loess. Along Mississippi River this may be as much as 10 feet thick, but at a distance of from 30 to 40 miles it becomes reduced to a few inches. Considering this erosion, one might conclude that few vertebrate remains would be preserved; nevertheless they are not wholly missing.

In 1862 (Geol. Surv. Wisconsin, vol. I, p. 136), J. D. Whitney stated that he had found in a crevice at Blue Mounds, Wisconsin, remains of the mastodon (p. 111), a peccary (p. 219), bones and teeth of a buffalo (p. 270), and a wolf which he referred with doubt to _Canis latrans_. The depth was uncertain, but it may have been as much as 40 feet. The fossils were embedded in reddish clayey loam, the usual crevice earth. On page 422 of the same volume, Jeffries Wyman referred the wolf remains to two distinct species, _Canis occidentalis_ and _C. latrans_. In 1876, Dr. J. A. Allen (Amer. Jour. Sci., ser. 3, vol. XI, pp. 47–49) described from the same lot of bones the species _C. mississippiensis_. This apparently did not include jaws and teeth that Wyman had referred to _C. occidentalis_. In Wyman’s paper, on page 422, he assigned three teeth to _Dicotyles torquatus_, an existing peccary, without stating that it had been found at Blue Mounds. In 1869 (Jour. Acad. Nat. Sci. Phila., ser. 2, vol. VII, p. 384), Leidy referred this peccary to his _Dicotyles lenis_, an extinct species. Inasmuch as the peccaries found at Galena were identified by Leidy (Whitney, vol. cit., p. 424) as _Platygonus compressus_ (p. 218), it appears pretty certain the _Dicotyles lenis_ (_Tagassu lenis_) was among the fossils collected at Blue Mounds (p. 219). It must, however, be kept in mind that Whitney, on page 35, stated that he had collected bones and teeth of the same animal near Dubuque, Iowa. Allen regarded the buffalo as belonging to an extinct species; but it is really undeterminable. Accordingly there may be credited to this locality the following species: _Tagassu lenis_, _Bison_ sp. indet., _Mammut americanum_, _Canis nubilus_ (_C. occidentalis_), _C. mississippiensis_, _C. latrans_.

In Whitney’s report, on page 133, he announced the finding of a large quantity of bones of mastodons at Sinsinawa Mound (p. 111), but he did not know at what depth they occurred. It seems probable that they had been met with in one or more crevices.

It seems probable that the animals found in crevices in the lead region of Wisconsin, Illinois, and Iowa belong approximately to one geological stage of the Pleistocene. The following appears to include all known to have occurred in such situations:

Megalonyx jeffersonii?. Platygonus compressus. Tagassu lenis. Odocoileus virginianus?. Cervus canadensis. C. whitneyi. Antilocapra americana. Bison sp. indet. Mammut americanum. Marmota monax. Microtus sp. indet. Geomys bursarius. Sylvilagus floridanus? (Lepus sylvaticus). Anomodon snyderi. Procyon priscus. Canis nubilus (C. occidentalis). C. mississippiensis. C. latrans.

The writer was at one time inclined to believe that these animals belonged to the time succeeding the withdrawal of the Wisconsin ice-sheet. Baker (“Life of the Pleistocene,” p. 353) thinks that they belong probably to the Peorian, inasmuch as the region is covered by Iowan loess, beneath which many of the bones have been found. It is quite probable that those crevices were open during at least some part of the Pleistocene and that animal remains collected in them. The fossils are reported as being sometimes inclosed in a matrix of cave or fissure materials which are cemented together by iron. The considerable number of extinct species, certainly 7 out of about 18, makes it probable that the fauna is not so recent as the Late Wisconsin.

It appears to be determined that the Iowan loess was formed immediately after the retirement of the Iowan ice-sheet. It might, therefore, be a question whether all of these animals might have got into those crevices in time to be covered in by the loess. On the other hand, the Illinoian drift was, for a long time, exposed to weathering and erosion before the Iowan drift and loess were laid down. Also, the Sangamon interval was probably much longer than the Peorian, so that the chances for the accumulation of the fossils were greater. It seems, however, that we can only say that the fossils are post-Illinoian and probably pre-Wisconsin.

Besides the vertebrate fossils referred to above, a few others, especially mastodons (pp. 110, 111), have been found at other places, but so little is known of the conditions of their interment that they furnish little geological information.

A very interesting region is found in the western part of the State, in Dunn and Pepin Counties. This has been examined with great care by Dr. Samuel Weidman, State geologist of Wisconsin. About Menomonie there are several brickyards, whose excavations furnish opportunities for studying the formations at that point. Sections of one of these brickyards are described and illustrated by Dr. E. R. Buckley, in Bulletin VII, part 1 (1901), page 194, plate XXXVII. A section and brief description is found also in a paper by Dr. Hussakof (Jour. Geol., vol. XXIV, p. 688). In that region are found outwash gravels which have been definitely correlated by Weidman with Iowan drift. In some places this is overlain by loess. These gravels vary from 10 to 20 feet in thickness at Menomonie. Beneath the gravels are found lacustrine clays varying in thickness from 20 to 40 and even 60 feet. These are stratified and consist of layers from 1 to 12 inches in thickness, with intervening thin layers of sand. Toward the bottom the sand increases in amount. Beneath the clay-bearing formation is a bed of sand attaining a maximum thickness of about 150 feet. This is underlain by coarse sand and gravel. The lacustrine clays and the underlying sands and gravels are included by Weidman in his Menomonie formation, and this is believed by him to be of Sangamon interglacial age. In northwestern Wisconsin are found other glacial deposits believed to belong to the Illinoian drift epoch.

In the lacustrine clay at Menomonie have been found remains of the great lake trout, _Cristivomer namaycush_ (Hussakof, as cited above), of a deer (p. 230), a caribou (p. 247), and probably a mastodon. The deer is represented by a single vertebra, identified by Dr. W. D. Matthew. The supposed mastodon is indicated by the distal end of the right femur, the caribou by an antler of a young and probably female individual and by the shaft of a large individual.

At Woodville, in St. Croix County, about 20 miles west of Menomonie, has been found a forest bed regarded as belonging to the Aftonian. This was described by Arthur Koehler (Amer. Forestry, vol. XXVI, Feb. 1916, p. 92, 3 figs.). Wood was found that was identified as that of spruce.

In 1913 (Science, n. s., vol. XXXVII, p. 457), in a brief abstract, Weidman reported that in Wisconsin he recognized drift deposits of Wisconsin, Iowan, and Kansan ages and another still older. No localities were mentioned, but his statements were doubtless based mostly on his work in the western part of the State. The loess was found to be laid down after the Iowan and before the Wisconsin. Interglacial deposits were found between the Kansan and the Iowan.

In 1905 (Jour. Geol., vol. VIII, pp. 238–256) and in 1910 (Jour. Geol. vol. XVIII, pp. 542–548), Dr. R. L. Chamberlain presented the results of his investigations on the “Pleistocene Geology of the St. Croix Region in Western Wisconsin.” His conclusion (p. 548) was that in that part of the State there were present (1) a surface mantle of gray Wisconsin drift deposited by a glacier from the Keewatin center; (2) red Wisconsin drift deposited by a glacier coming from the Labrador center; (3) a red drift left by an ice invasion from the Labrador center, its age consistent with Illinoian; (4) a grayish-black till that had come from the Keewatin center and whose age was probably Kansan.

MARYLAND AND DISTRICT OF COLUMBIA.

For obvious reasons the Pleistocene geology of the District of Columbia is considered in connection with that of Maryland. This region is of especial interest, because of the long time and the care which has been bestowed on it by geologists and because the conclusions reached have been applied to the geological study of States both toward the north and toward the south.

The most complete exposition of the Pleistocene geology of the region is to be found in the volume of the Maryland Geological Survey entitled “Pliocene and Pleistocene,” published in 1906. The geological treatise itself was written by George Burbank Shattuck and is illustrated by many maps and text-figures. Included in this is a bibliography of the subject which occupies 17 pages. There is a chapter by W. B. Clark, Arthur Hollick, and F. A. Lucas, on the interpretation of the palæontological criteria; another by F. A. Lucas on the mastodons and the elephants. The Pleistocene mollusks found in the State, 40 species, were described and figured by W. B. Clark; while the plants, also nearly 40 in number, were described and figured by Arthur Hollick.

The history of the development of our present knowledge of the geology of Maryland and the classification of its formations up to 1906 is given by Shattuck in the volume just cited (pp. 25–40). This geologist recognized in the superficial deposits of the State five formations (fig. 15). These are, beginning with the oldest, Lafayette, Sunderland, Wicomico, Talbot, and Recent.

The Lafayette is regarded as having been laid down during the Pliocene. The Sunderland, Wicomico, and Talbot form three terraces, of which the Sunderland is the oldest, most elevated, and farthest away from the larger bodies of water. It is composed of clay, peat, gravel, and boulders supposed to have been brought in by the ice. The coarser materials appear to occupy usually the lower parts of the formation. The elevation near Washington is about 200 feet, but southward it descends gently, until in St. Mary’s County it is only about 60 feet. The thickness varies from about 80 feet to nothing. According to Shattuck, at the time of deposition of the Sunderland the coast was depressed to an extent of about 200 feet, so that its materials were laid down either in salt water or in that of wide estuaries. No deposits belonging to it have been found in the eastern peninsula. In the western peninsula considerable areas are recognized along the Potomac up to Washington and along the Patuxent and Chesapeake Bay to Baltimore and Elkton. Except in the southern part of this peninsula, the Sunderland is found only in widely separated patches. No marine organisms are known to have left their remains in the Sunderland, but forest trees of a number of existing genera and several extinct species have been described by Hollick in the volume cited.

The Wicomico formation is described as occupying a large portion of the central and higher parts of the eastern peninsula; in the western it forms a narrow and often interrupted fringe around the Sunderland. North of Washington and Annapolis it occurs only in patches. Its materials are very similar to those of the Sunderland. Its greatest elevation is about 100 feet above sea-level, and this, according to Shattuck’s view, marks the amount of depression of the land at that time. The thickness may be as much as 70 feet, but is usually much less. No marine fossils proper to the period have been discovered in the deposits, but at a point in Prince George’s County plant remains have been found in a deposit about 20 feet thick.

The Talbot formation forms a fringe, sometimes of great width, sometimes narrow or interrupted, along all the large bodies of water in this State and in Delaware. It is the lowest of the terraces. The greatest elevation is about 45 feet; the thickness does not exceed 40 feet. The materials noted are those of the other two formations—clay, peat, sand, gravel, and ice-borne boulders. At several points along Chesapeake Bay and on the lower part of Patuxent and Potomac rivers, deposits containing plant remains have been discovered, including pines, cypress, hickory, beech, elm, and black locust. In contrast with the other formations, the Talbot has furnished many marine fossils, mostly mollusks; but in all cases the localities are close to the present coast.

The writer does not accept the theory that the materials forming what have been called the Sunderland, Wicomico, and Talbot terraces have been to any great extent laid down in the sea. Some part of the Talbot, that lying near the present coast, has undoubtedly had such an origin. Nor has the Coastal Plain suffered, so far as is determinable, any such amount of depression as the theory mentioned requires. The materials of the Sunderland and Wicomico have, in the writer’s opinion, been brought down by rivers whose beds lay at levels nearly as high as those of the real or supposed terraces. When the Talbot materials were laid down, the rivers and estuaries of the coast had been cut down nearly to their present levels, and this was not long after the beginning of the Pleistocene.

The authors of the submergence theory admit that no satisfactory evidence of the presence of marine organisms, vertebrate or invertebrate, are to be found in the body of the assumed terraces, except again in parts of the Talbot which immediately border the ocean or the great estuaries. It is almost inconceivable that the ocean could occupy the Coastal Plain from New Jersey to Mexico for thousands of years and lay down great thicknesses of clay, sand, and gravel without having left somewhere beds of molluscan shells in such situations that they would have been discovered. While these marine fossils are lacking, there are found on all these terraces from Maryland to Florida and to the Rio Grande an abundance of land vertebrates such as elephants, mastodons, horses, camels, peccaries, and many other forms. Nor do our palæobotanists have difficulty in finding oaks, walnuts, hickories, poplars, etc. On the theory of submergence there are missing all the things that ought to be found and there are met with just the things that would not be expected.

A figure is here reproduced (fig. 15) from the Maryland Pliocene and Pleistocene volume, page 66, with the explanation there accompanying it. The reader may judge for himself whether the sea could occupy the Atlantic coast since Pliocene times without leaving any traces of marine fossils, while at the same time there were preserved in those terraces remains of land animals and land vegetation.

Another section (fig. 16) is reproduced from Folio 179 of the U. S. Geological Survey, the authors of which are G. W. Stose and C. K. Swartz. The uppermost terraces are by these authors supposed to belong to the late Pliocene, the formation formerly known as the Lafayette. These figures suggest that the one set of terraces have some connections with the other set.

Beginning at the southern extremity of Maryland, we notice the occurrence of remains of _Mammut americanum_ at or near St. Mary’s City. Other remains of the same animal have been secured near St. Clements in St. Mary’s County (p. 112). Both of the localities are situated on territory mapped by Shattuck as Wicomico; but as remarked on page 112, our knowledge of the conditions under which the fossils were found is not sufficient to allow us to say more than that they belong to the Pleistocene. The species existed from early to late Pleistocene and can not be used to determine the age of the deposits.

Along Patuxent River, in Charles County, not far from Benedict, Cope (Proc. Acad. Nat. Sci. Phila., 1867, p. 155) recognized jaws and teeth of _Grison macrodon_ and of _Tagassu lenis_ (p. 220). Both are extinct species.

According to Shattuck’s map of 1906, this region is covered by the Talbot formation; but inasmuch as the species named were obtained from pits furnishing Miocene marl, one can not be sure that they are not older than the supposed Talbot. It would probably require a search in the land records in order to determine exactly where the objects were found. The presence of _Elephas primigenius_ suggests that this animal had been pushed down here during one of the glacial stages.

Nearly a hundred years ago an elephant tooth (p. 154) was found somewhere in Queen Anne County, but it would probably be now impossible to determine the locality. In case the elephant tooth was found near Chesapeake Bay, as is very probable, there is no record of any Pleistocene vertebrate having been found in the central and eastern parts of the eastern peninsula.

In the eastern peninsula remains of Pleistocene vertebrates have been recorded from only two localities, Oxford Neck, Talbot County, and an undetermined locality in Queen Anne County. From Oxford Neck, Cope (Proc. Amer. Philos. Soc., vol. XI, 1869, p. 178) reported _Elephas primigenius_, _E. columbi_, _Cervus canadensis_, _Odocoileus virginianus_, _Chelydra serpentina_, and _Terrapene eurypygia_.

At Chesapeake Beach, William Palmer, of the U. S. National Museum, discovered a few remains of Pleistocene vertebrates. One of these is a tooth of an undetermined species of _Bison_, probably not the existing one. Another species is probably _Equus leidyi_ (p. 189). Three teeth appear to represent the peccary _Tagassu lenis_ (p. 220). In 1921, Dr. Adolph H. Schultz, of the Johns Hopkins Medical School, presented to the U. S. National Museum another specimen of _T. lenis_ which he had found at Chesapeake Beach. Inasmuch as the fossils were picked up after having fallen from their resting-place, it is impossible to say to which formation they belonged. In the opinion of the writer, none of the three species indicates a late Pleistocene time.

On the opposite side of the western peninsula, at Marshall Hall, Charles County, there was found long ago a tooth which the writer refers to _Equus leidyi_.

Coming north into the District of Columbia, we find recorded the discovery of remains of horses and possibly at two different times. According to Darton’s work (Folio 70, U. S. Geol. Surv.), there is some later Columbia laid down along the route of the Chesapeake and Potomac Canal above Georgetown. This would now doubtless be regarded as belonging to the Talbot. It seems to follow that either the Talbot is much older than has been supposed or that some of the extinct horses continued on until a comparatively late time in the Pleistocene.

Within the limits of the city of Washington there has been found a tooth of probably _Elephas primigenius_ at a depth of 35 feet, in the Wicomico formation (see p. 178). On any theory of the origin of the terraces, the presence of the tooth at that depth in the ground and at that elevation appears to indicate a considerable geological age for the animal. To what extent materials may have been washed down from the surrounding higher land may be difficult to determine.

In Prince George County, near Mitchellville, have been found two teeth of an extinct horse (p. 188). These are as yet unidentified. They are in the U. S. National Museum, No. 8813.

Near Towson, in Baltimore County, a mastodon tooth has been found (p. 112); but beyond proving that there is at that locality some Pleistocene deposit, it gives us little information.

In 1920 (Proc. U. S. Nat. Mus., vol. LVIII, pp. 96–109), the writer described a collection of vertebrate fossils, collected in a cave or fissure in limestone at Cavetown, Washington County, by anthropologists from Phillips Academy, Andover, Massachusetts. The following is the list of species that were found in the collection:

Crotalus horridus. *Equus complicatus (p. 189). *Equus giganteus? (p. 189). *Mylohyus nasutus (p. 220). *M. exortivus (p. 220). *M. obtusidens, n. sp. (p. 220). *Platygonus tetragonus? (p. 220). *P. vetus? (p. 220). *P. cumberlandensis (p. 220). *Sangamona fugitiva. Odocoileus virginianus (p. 231). *Elephas columbi? *Sciurus tenuidens, n. sp. S. hudsonicus. S. carolinensis. Marmota monax. Castor canadensis. Ondatra zibethica. Neotoma magister. Microtus pennsylvanicus. Erethizon dorsatum. Sylvilagus floridanus. Ursus americanus. *Smilodontopsis mooreheadi. Felis couguar.

Of the 22 species here recognized 12 are extinct. This large number of itself indicates that their time of existence was not recent. Similarly, the presence of 2 species of horses, several species of peccaries, and of a saber-tooth tiger points to a rather ancient period. The writer believes that the assemblage belongs to the Sangamon stage of the Pleistocene.

In Washington County, probably along Lane’s Creek, was found, in digging a mill-race, the skull of a mastodon (p. 112). Further east, near Clear Spring, and about a mile above the entrance of Conococheague Creek into the Potomac, was discovered a tooth of a mastodon (p. 113). This had been washed out of some deposit along this creek, probably not far away from where it was found. As Stose has shown (Hancock Folio, No. 179, U. S. Geol. Surv.), along the Potomac and its tributary streams there are extensive Pleistocene deposits of sand and gravel, laid down when the river was as much as 200 feet above its present level. It is probable that such deposits date from the early Pleistocene (fig. 17). A more important locality for Pleistocene vertebrates is that near Corriganville, about 3 miles west of north of Cumberland, Maryland. The cave is in Allegany County, west of Wills Creek and south of Jennings Run, about 0.5 mile south of the village of Corriganville. An account of this locality, with a list of the species determined up to that time, has been published by Gidley (Proc. U. S. Nat. Mus., vol. XLVI, 1913, pp. 93–102). In cutting through a spur of limestone in making a railroad, at a depth of about 100 feet there was exposed a cave or fissure which contained many bones and teeth. Gidley secured some hundreds of specimens belonging to about 35 species. Unfortunately nothing has been published which shows the relation of this cave to the terraces which are found along Potomac River and its tributaries. Through the kind offices of Mr. F. S. Rowe, welfare agent of the Western Maryland Railway, the writer has received from the division engineer, Mr. P. Cain, of Cumberland, a topographic map of Allegany County and a profile of the road extending through the rock cut. From these it appears that the level of the track, at the fissure, is 837 feet above sea-level. This seems, therefore, to be considerably above the highest terrace along the Potomac in that region. It is to be supposed that the fissure was formed long before the animal remains accumulated in it.

In a paper published in 1920 (Proc. U. S. Nat. Mus., vol. LVII, pp. 651–678, plates LIV, LV, text-figs. 1–10) Gidley added to his former list four species of peccaries, as follows: _Platygonus cumberlandensis_, _P. intermedius_, _Mylohyus exortivus_ (all new), and _M. pennsylvanicus_. In another communication he reported also a deer, a wolverine, a beaver, a lynx, a badger, a marten, an eland, and a crocodile or an alligator (Rep. Smithson. Inst. for 1918, pp. 281–287). Many of the identifications are merely provisional.

_Provisional list of fossils found near Corriganville._

1. Alligator or Crocodylus sp. indet.

2. Blarina brevicauda?.

3. Vespertilio grandis.

4. Vespertilio sp. indet.

5. Myotis sp. indet.

6. Ursus vitabilis.

7. Ursus americanus?.

8. Canis armbrusteri.

9. Canis sp. indet.

10. Vulpes? sp. indet.

11. Mustela vison?.

12. Gulo luscus?.

13. Taxidea sp. indet.

14. Lynx sp. indet.

15. Mammut americanum.

16. Equus sp. indet. (p. 189).

17. Tapirus haysii? (p. 204).

18. Platygonus cumberlandensis (p. 220)

19. P. intermedius (p. 220).

20. P. vetus? (p. 220).

21. Mylohyus exortivus (p. 220).

22. M. pennsylvanicus (p. 220).

23. Odocoileus sp. indet.

24. Taurotragus americanus.

25. Ochotona princeps?.

26. Lepus americanus?.

27. Lepus sp. indet.

28. Sciurus hudsonicus.

29. Sciuropterus alpinus?.

30. Marmota monax?.

31. Castor sp. indet.

32. Neotoma sp. indet.

33. Microtus chrotorrhinus?.

34. Synaptomys borealis?.

35. Synaptomys sp. indet.

36. Peromyscus leucopus?.

37. Napæozapus sp. indet.

38. Erethizon sp. nov.

On account of the present unstudied condition of the collection, it is difficult to reach conclusions that are satisfactory. It appears, however, that there are at least 6 hitherto undescribed species, one-fifth of the whole number. Another 6, if at all correctly determined, indicate a wide removal from their ranges of the present day. _Lepus americanus_ now lives well toward the north, coming down to Saginaw, Michigan. _Ochotona princeps_ lives in the Rocky Mountains of British America. _Synaptomys borealis_ is known only from the region about Great Bear Lake, Mackenzie, Canada. _Microtus chrotorrhinus_ has its habitat in Quebec and the northeastern United States. The species of _Napæozapus_ are Canadian in their range, but descend to southeastern Maryland and to North Carolina in the mountains. _Sciuropterus alpinus_ is found from Alaska to Hudson Bay, but descends on the Pacific coast to southern California. This northern habitat of so many supposed species suggests that the fissure received its contents during one of the glacial stages, and this may be the case. However, it is not unlikely that these species and some others are really undescribed ones. One may reasonably expect to find in a fauna containing _Equus_ and _Tapirus_ a much higher percentage of extinct species than Gidley has recorded.

The most remarkable member of the fauna is _Taurotragus americanus_, a species closely related to the eland of southern Africa (Gidley, Smiths. Misc. Coll., vol. LX, No. 27). Its presence in western Maryland gives a vivid impression of the widely extended journey that some animals have made from one continent to others. The same species has since been found in collections made at Alton, Illinois (p. 339), and at Kimmswick, Missouri (Proc. U. S. Nat. Mus., vol. LVIII, p. 113).

According to the author’s views, the fauna found at Cumberland, like that of localities in western Virginia, belongs to a time somewhere about the middle of the Pleistocene. Most of the species may be supposed to have lived there during the warm Sangamon stage; others, as the wolverine, at a somewhat earlier or later time when the climate was cooler.

VIRGINIA.

For the student of Pleistocene vertebrate palæontology, as for the geologist, Virginia may be divided into three physiographic regions, the Coastal Plain, the Piedmont Plateau, and the Appalachian Mountains. The line which divides the Coastal Plain from the Piedmont Plateau begins at the southern boundary of the State, at about 77° 31′ longitude. The towns on or not far from this nearly north-and-south line are Emporia, Petersburg, Richmond, Hanover, and Fredericksburg. Near the latter the line inclines slightly eastward and passes a few miles west of Alexandria and Washington, D. C. The Coastal Plain is much less elevated than the region west of it and consists of deposits of Mesozoic or Cenozoic age, and much of it is covered by Pleistocene materials. The Plateau region is elevated and consists mostly of Palæozoic rocks, mostly metamorphosed into a crystalline condition. The Appalachian region presents nearly parallel ranges of mountains and intervening valleys.

For a knowledge of the Pleistocene geology of the Coastal Plain the reader should consult Bulletin iv, 1912, of the Virginia Geological Survey. The authors who discuss the physiography and geology of this region are William B. Clark and Benjamin L. Miller. On pages 19 to 45 they present a very full bibliography of the geological literature pertaining to this region. Additional valuable assistance may be obtained from the various folios issued by the United States Geological Survey, but unfortunately not many species of vertebrate animals have been found on this Coastal Plain of Virginia.

In Bulletin IV, already mentioned, Clark and Miller recognize the presence of three terraces belonging to the Pleistocene. To these are given the names applied in Maryland and North Carolina to what are regarded as equivalent terraces. The oldest of these, most elevated and farthest from the coast, is the Sunderland; eastward of this lies the Wicomico; the Talbot is the youngest and lowest and borders the coast. Unfortunately, the geologists referred to did not map the areas occupied individually or collectively by these terraces. They accept the theory that these terraces were laid down in the sea. It is admitted, nevertheless, that no marine fossils are found in deposits of the Sunderland and Wicomico. In the Talbot, 26 species of marine mollusks have been reported from Talbot deposits of the Dismal Swamp Canal, all regarded as belonging to living species. It will be recollected that Woolman (Proc. Acad. Nat. Sci. Phila., 1898, p. 414), in a study of mollusks collected in the Dismal Swamp Canal, found 7 extinct species in a collection of 49 species, equal to about 16 per cent. It is, however, not unlikely that the collections had been dredged up from deeper deposits.

In Bulletin V of the Virginia Geological Survey, on page 25, Sanford stated that the Talbot had a width of 30 miles at the south. On consulting Stephenson’s map of the superficial formations of the Coastal Plain in North Carolina (North Carolina Geol. Surv., vol. III, plate XIII) it will be seen that this corresponds quite exactly with the width of the Pamlico formation at that line. For the writer’s views on the terraces named the reader may consult page 346 on the geology of Maryland.

On page 113 is recorded the discovery of a tooth of a mastodon in a marsh near Disputanta, in Prince George’s County. Not enough is known about the geology of the region to say more than that the deposit belongs to the Pleistocene.

About 6 miles east of Williamsburg, a little more than 100 years ago, remains which pretty certainly belonged to the genus _Mammut_ and probably to the species _M. americanum_ (p. 113) were discovered, said to have been found on the banks of York River; but by this was probably meant the banks of the flood-plain. The bones were found in marsh mud and were surrounded by roots of cypress trees. The adjacent bank was 20 feet higher than this level. The topographical map of the Williamsburg Quadrangle shows that an abrupt rise of this amount is to be found only about 10 miles away from the river. Whether the cypress roots were those of trees that had grown within recent years or whether they were remains of a Pleistocene forest, such as was exposed at Tappahannock, Essex County (Bull. IV, p. 186), the writer does not know. The information at hand about this case does not make it possible to pronounce on the geological age of the mastodon.

On page 28 an account is given of the discovery of a skull of a walrus on the Atlantic coast of Virginia, at Accomac. It had doubtless been washed up by the sea from a Pleistocene deposit. It is easiest to suppose that the walrus had been driven southward along the coast during the Wisconsin glacial stage; but possibly this happened during an earlier glacial time.

No vertebrate fossils of Pleistocene age appear to have come to light anywhere on the Piedmont Plateau, and little or nothing is known about its Pleistocene geology.

From the geological surveys we get little information about the Pleistocene formations of the Appalachian region. At most, mention is made of soils of undetermined age along the streams; and yet from this region have been obtained a very considerable number of Pleistocene vertebrates.

From Mr. Wyndham Robinson, of Abingdon, Washington County, the U. S. National Museum received in 1869 a tooth of _Mammut americanum_ (p. 113) and one of _Equus complicatus_ (p. 189). Nothing has been learned regarding the conditions under which they were unearthed. The horse-tooth points to an age preceding the Wisconsin drift.

From Saltville, in Smyth County, the following forms have been obtained:

Crocodylus sp. indet. Megalonyx dissimilis (p. 34). Equus sp. indet. (p. 190). Odocoileus? sp. indet. (p. 231). Cervalces sp. indet. Bison sp. indet. (p. 259). Mammut americanum (p. 113). Elephas primigenius (p. 145).

That a crocodile should have lived in this region during the Pleistocene is remarkable. _Megalonyx dissimilis_ is otherwise known only from Natchez, Mississippi, from deposits which appear to be of about Illinoian or Sangamon age. The horse-tooth points to about this time or earlier, while the other species do not contradict this conclusion. The astragalus referred to _Odocoileus_ probably belongs to some other genus.

Mr. M. D. Mount sent to the U. S. National Museum remains of _Bison_ (p. 259), _Mammut americanum_ (p. 113), and _Elephas primigenius_ (p. 145). These, he reported, had been found at a depth not greater than 8 feet in excavating for the city reservoir. He has written that the valley of Holston River at Saltville, within about 80 years, had been a lake, at least at certain times of the year, and that the reservoir was excavated at the margin of this low area.

Mr. O. A. Peterson (Ann. Carnegie Mus., vol. XI, 1917, pp. 469–474) reported from this place the crocodile, the megalonyx, cervalces, the supposed deer, the horse-tooth, and remains of mastodons. The bones were found in a sink-hole, in a layer of coarse gravel, pebbles and cobblestones, a fact indicating that a stream of some size had occupied the place. Overlying this layer was one in which there were fragments of large river shells. The bone layer appears to have been only about 4 feet from the surface. Peterson concluded that at the close of the Pleistocene or later the remains had been moved and redeposited from some place not far away, but this would not affect the geological age of the fossils and it is evident that remains of vertebrates are widely dispersed in that valley. All the species reported are extinct, but only large forms were secured.

Professor Cope, probably in 1868, found the following 24 species. He did not state the localities exactly, except that they were along New River, in Wythe County. Two were on the land of Abraham Painter. The writer applied to the surveyor of the county named and has been informed that the farm which belonged to Abraham Painter is on New River, near the town of Ivanhoe. The nomenclature of the species has been revised. The species preceded by a dagger are extinct.

†Megalonyx jeffersonii (p. 34). Castor fiber. Neotoma floridana? Marmota monax. Peromyscus leucopus. †Tamias lævidens. †Sciurus panolius. Sylvilagus floridanus. Blarina sp. indet. Vespertilio sp. indet. †Tapirus haysii (p. 204). †Equus complicatus? (p. 190). †Mylohyus nasutus (p. 221). Odocoileus virginianus (p. 231). †Bison sp. indet. (p. 260). †Ursus amplidens. Procyon lotor. Spilogale putorius. †Myxophagus spelæus. Crotalus sp. indet. Amyda sp. indet. Terrapene sp. indet. Cryptobranchus sp. indet.

At least 9 of the 24 species are extinct. None of the recorded species requires us to refer the deposit to early Pleistocene times. _Ursus amplidens_ was described from the deposits at Natchez. This and _Tapirus haysii_, _Equus complicatus_, and _Mylohyus nasutus_ point to middle Pleistocene, apparently about to Illinoian or Sangamon times.

Cope reported that the teeth and bones were found in a cave breccia. This consisted of a number of irregular masses which occupied “depressions and short galleries” in the southeast side of a line of hills. When those masses were excavated from their beds the floor and roof of a portion of a cave were exposed, with the stalactites, stalagmites, and usual incrustations. It would appear, therefore, that at some time in the early Pleistocene or in the late Pliocene the caves had been formed through the effect of streams of carbonated waters on the limestone; that in some way the bones and teeth of the animals listed above had got into the cave; that by a change in the amount or character of the water the caves had gradually filled up; and that afterwards the limestone which contained these caves had undergone great erosion.

Further north, in the valley of Jackson River at Covington, there is evidently a deposit of Pleistocene clay, for in it at a depth of 12 feet was found a tooth of a mastodon (p. 114). Another mastodon tooth was found near Hot Springs, at the head of Wilson Creek, in Bath County, possibly in similar deposits (p. 114). In Augusta County an unidentified species of horse (p. 190) and the peccary _Platygonus_ (p. 221) have been discovered.

WEST VIRGINIA.

So far as the writer has learned, vertebrate remains belonging to the Pleistocene have been found in West Virginia in only eight places and only seven species are represented: _Mammut americanum_ (p. 115), _Elephas_ sp. indet. (p. 179), _Equus niobrarensis?_ (p. 190), _Symbos cavifrons_ (p. 254), _Megalonyx jeffersonii_ (p. 34), _Odocoileus virginianus?_ (p. 231), and a peccary (p. 221). The horse appears to indicate an early Pleistocene time, possibly pre-Kansan, but all the other species continued from at least the Aftonian stage through to the Late Wisconsin. The specimens, therefore, do not help us to determine the age of the deposits in which they are found.

No part of the State lies within the glaciated area; hence, during the whole of the Pleistocene epoch its surface was subjected to weathering and to the erosion of running water. At times the streams built up deposits on their beds. Later they deepened their channels and left a part of their former deposits as terraces. At a still later time the deposition and deepening may have been repeated, and as a result there is sometimes a series of terraces one above another. The age of these terraces and their origin have been the subjects of a good deal of controversy.

In the Masontown-Uniontown Folio (U. S. Geol. Surv. No. 82), M. R. Campbell has discussed the terraces along the Monongahela River, which occur at an altitude of about 1,000 feet above sea-level and perhaps 150 feet above the present river. Also more than 100 feet above the present river are old abandoned river channels which are now partially filled up.

In 1911 (U. S. Geol. Surv. Folio 178, pp. 11–13), E. W. Shaw and M. J. Munn described the Quaternary of the Foxburg and Clarion quadrangles in Pennsylvania, where the same Pleistocene problems are involved. They present an account of the different views regarding the high-level terraces and the abandoned channels. They concluded, as did Campbell, that these terraces and channels dated back to the early Pleistocene and probably to the Kansan stage. Figure 17 is a reproduction of Shaw and Munn’s figure 10, on their page 12. It represents a section across Allegheny River at Parker’s Landing, Armstrong County, Pennsylvania. The uppermost gravels in the figure would be those of supposed Kansan age; while the lowermost are those laid down during the last glacial stage, the Wisconsin. In the materials of the high terraces one may expect to find fossil vertebrates of the early Pleistocene, as in the case of the mastodon reported from Stewartstown, West Virginia (p. 116). The conditions of burial should, however, be carefully studied and recorded; for it would be possible for remains to be left at a later time on such a terrace and to be covered up by earth washed down from above.

On page 254 an account is given of finding a musk-ox skull near Steubenville, Ohio, on a terrace about 75 feet above the low-water mark. The region of the western part of West Virginia, western Pennsylvania, and northeastern Ohio is interesting because of its history during the late Pleistocene. The reader is referred to Leverett’s monograph, “The Glacial Formations and Drainage Features of the Erie and Ohio Basins” (Monogr. U. S. Geol. Surv., vol. XLI, 1902, pp. 88–158, with figs.). Leverett essays to show that the upper part of the Ohio River, the Allegheny, and the Monongahela with its branches at one time emptied into Lake Erie. The connection was made through Beaver River, which now flows into the Ohio, and Grand River, in eastern Ohio, now emptying into Lake Erie. When the Wisconsin ice filled Lake Erie and occupied its southern shore the mouth of Grand River was dammed and the water could escape only to the south. The flow was reversed, and after it had reached the top of the divide it entered the stream that then represented the head of the Ohio. When at length the mouth of Grand River was reopened, the new channel had been cut so deep that most of the streams of western Pennsylvania and West Virginia continued to flow down the Ohio. Leverett’s figure representing the preglacial drainage of the upper Ohio region is here reproduced (fig. 10).

NORTH CAROLINA.

(Map 39.)

Our knowledge of the Pleistocene geology of North Carolina is at present confined almost wholly to the Coastal Plain of the State. The most recent general discussions of the geology of this region are found in