volume III of the North Carolina Geological and Economic Survey, 1912.

The authors who contributed to this volume are William Bullock Clark, Benjamin L. Miller, L. W. Stephenson, B. L. Johnson, and Horatio N. Parker. L. W. Stephenson has furnished an article on the Cretaceous deposits, and in his numerous geological sections he has referred to the Pleistocene materials there found. Benjamin L. Miller wrote on the Tertiary formations and likewise noted the Pleistocene materials found in his sections. The most important part of the volume for the student of the Pleistocene is Stephenson’s article on “The Quaternary Formations,” which occupies pages 266 to 290. Clark, Miller, and Stephenson united in a chapter on the “Geological History of the Coastal Plain of North Carolina.” Clark, besides, deals with the “Correlation of the Coastal Plain Formations.” In addition to numerous plates and text-figures, a colored map shows the area covered by the surficial formations of the Coastal Plain and another the distribution of the formations exclusive of the surficial. Finally, Miller and Stephenson presented a bibliography which includes 150 titles, occupying pages 44 to 73.

According to Clark and Stephenson, the Pleistocene of North Carolina comprises five formations; the oldest is the Coharie, farthest removed from the coast and lying back against the so-called Lafayette, itself supposed, with some doubt, to belong to the Pliocene. Toward the coast there come in, in succession of position and time, the Sunderland, the Wicomico, the Chowan, and the Pamlico. These formations are described as forming more or less well-defined terraces having higher and higher elevations as they are followed back from the coast. The Pamlico nowhere exceeds 25 feet above sea-level. The Chowan varies in elevation from about 25 feet to about 50 feet. The Wicomico formation slopes from about 50 feet up to about 90 or 100 feet. The Wicomico may attain elevations of from 140 to 150 feet at the western border. The Coharie varies from about 160 feet along its eastern border to as much as 235 feet along its western border. From its western border each formation sends up the rivers prolongations into or across the next formation toward the west.

Each terrace may present along its coastward border an escarpment of varying elevation and obviousness. The Coharie and Sunderland formations are regarded by the authors named as being correlated with the Sunderland of Virginia and Maryland, although the Coharie may be really Pliocene. The Wicomico is equivalent to that called by the same name in the States farther north, while the Chowan and the Pamlico together are correlated with the Talbot of Virginia and Maryland.

The area occupied by the Pamlico is extremely narrow or absent along the southernmost third of the coast of the State. At longitude 77° the boundary between it and the Chowan turns and runs north, very slightly to the east, striking the northern boundary of the State at about 76° 15′. Just south of Albemarle Sound its width east and west is nearly equal to that of all the other Pleistocene formations at that latitude, taken together.

Clark, Miller, and Stephenson (op. cit., p. 300) accept the theory of McGee that during Lafayette times, probably in the late Pliocene, the Coastal Plain was depressed some 500 feet below its present level and covered by the sea. Into this sea were poured, by the rivers coming down from the higher lands to the west, the clay, sand, and gravel, sometimes boulders, which make up the so-called Lafayette. Somewhat later the region was uplifted enough to expose the Lafayette deposits and they suffered erosion. When the Coharie formation began to be laid down the sea-level must have been about 160 feet higher than at present; it continued to rise until it reached an elevation of about 200 feet. A subsidence and a succeeding elevation occurred, during which the Sunderland terrace was produced. In like manner the succeeding deposits and terraces are supposed, by the geologists named, to have been formed—the Wicomico, the Chowan, and the Pamlico.

One objection already offered (p. 346) to this theory to account for the deposits belonging to the Lafayette and the formations of the Pleistocene is that, instead of beds of sea-shells, remains of marine fishes, porpoises, and whales, there are found scattered here and there over this region the bones and teeth of elephants, mastodons, horses, and other land animals. In maintaining this objection it is not necessary to assume that the lower parts of the Pleistocene area have never been submerged.

The writer has caused to be prepared a map showing the geographical distribution of the five formations referred (in the work cited) to the Pleistocene. It is based on the maps found in that volume. It shows also the localities where fossil vertebrates have been discovered, and where marine fossils and land plants have been secured (map 39).

One difficulty met with in our study of the distribution of the finds of extinct vertebrates in North Carolina, as elsewhere, arises from carelessness in recording and preserving proper data. In several cases here to be considered, no more is known than that a fossil has been found in a certain county. Happily, more is known in many other cases.

Examination shows that no fossil vertebrates are known to have been found in North Carolina within the area of the Coharie formation, but that mastodons have been met with in the areas of all four of the other formations as laid down in Stephenson’s map, plate XIII of the work cited above. Horse remains, too, seem to have occurred within all the areas last noted. This does not mean necessarily that these remains were buried in the corresponding formations. A mastodon may have lived long after the Sunderland was laid down and his remains have become buried in some isolated deposit, say of Pamlico times; or, the remains may be found within the area of Pamlico, but really buried in underlying Chowan. Each case must be decided on the evidence bearing on it.

Mention is made on page 155 of the finding of a tooth of _Elephas columbi_ about 9 miles below Wilmington. Whether this was buried in Pamlico deposits close along Cape Fear River, in Chowan deposits which prevail there, or beneath these, in Wicomico, it is impossible to say. A short distance below this place was found a tooth of _Mammut americanum_.

On page 190 is given an account of the discovery of a tooth of _Equus leidyi_ in what was supposed to be Miocene marl in the vicinity of Elizabethtown, on Cape Fear River, in Bladen County. Miller (op. cit., p. 248) states that the Pleistocene about Elizabethtown rests usually directly on the Cretaceous, but that south of the town are found some patches of Miocene marls. The region about this town is mostly occupied by the Sunderland formation, but the Wicomico extends up the river far above the place. It is, however, mapped as lying mostly on the north side of the river. It seems pretty certain that the horse-tooth occurred in the Sunderland, probably at its base.

Mastodon remains, as stated on page 115, have been found in Pender County, but where is not known. Along the coast is a narrow strip of Pamlico. The southeastern half of the county is occupied by the Chowan, the northwestern by the Wicomico.

Mastodon teeth have been found in Duplin County, but there is no record as to exact locality, depth, or matrix. The southeastern two-thirds of the county is covered by deposits of the Wicomico, the northwestern third by Sunderland. The mastodon probably belongs to one or the other of these. The Pleistocene deposits are, however, underlain by Tertiary rocks, and possibly the mastodon came from these and belongs to a different genus.

On page 116 will be found an account of remains of a mastodon, probably _Mammut americanum_, which was found near Jacksonville, in Onslow County. Three of the supposed Pleistocene formations are found near Jacksonville. The Pamlico comes up the New River quite to the town. Immediately at the town is (following Stephenson’s map) the Chowan. The southeastern border of the Wicomico comes down nearly to the town. In which of the three areas the teeth were discovered we do not know. A case is here furnished which illustrates the need of most accurate observation and record of locality, depth, and character of materials.

As stated on page 116, teeth and tusks of _Mammut americanum_ have been obtained at Maysville, Jones County. The writer does not know exactly the place where the remains were discovered. The region about Maysville is occupied by the Chowan formation, but the Pamlico sends an extension up White Oak River as far as Maysville.

Remains of both _Mammut_ and _Elephas_ have been reported from Carteret County. In 1828 (see p. 117) Elisha Mitchell stated that remains of the elephant and mastodon had been met with in digging the Clubfoot and Harlow Canal. This canal passed from Neuse River to Newport River. In 1876 (Senate Ex. Doc. No. 35, 44th Congr., p. 17) S. T. Abert transcribed, from an earlier report made by Professor Olmstead, a geological section taken in this canal. The excavation went to a depth of 16 feet. The uppermost of the four layers consisted of the peaty mold usually found in the swamp. The next layer was made up of a yellowish-brown potter’s clay. The third layer consisted of sand and was full of sea-shells and fossil remains of “mammoths” (mastodons) and elephants. The shells belonged to species now found near Cape Lookout, principally conch, scallop, and clam. The layer below this was blue clay. In the case here presented there can hardly be a doubt that the stratum containing the shells and the bones belonged to a Pleistocene formation older than that assigned to the Pamlico.

On page 145 is described a tooth of _Elephas primigenius_, dredged up in Core Creek, forming part of the Inland Waterway in Carteret County. The conclusion seems unavoidable that this boreal animal had been driven to this southern latitude during one of the glacial stages, and one naturally thinks of the latest one, the Wisconsin; but it may have been at a much earlier time. A mastodon jaw has been secured in the same canal.

Doubtless the locality in North Carolina, the most important to the student of Pleistocene vertebrate palæontology, is that reported long ago on the northern shore of Neuse River, 16 miles below Newbern. As stated on page 117, in a mention of the mastodon bones discovered, H. B. Croom seems first to publish a statement concerning the animal remains found there. Some of his identifications were certainly wrong. According to Harlan (Amer. Jour. Sci., vol. XLIII, 1842, p. 143), there were secured remains of elephant, mastodon, hog, elk, deer, horse, seal, cetaceans, a tortoise, snake, fish, shark, and skate. As in another case, Harlan may have mistaken worn teeth of _Bison_ for teeth of the hog (_Sus_). For our purpose the most important animals of the list are the elephant, the mastodon, and the horse. According to Croom, the animal remains were found in a marl pit. He was informed by the owner that in an upper layer there were found teeth of sharks and fragments of bones of marine fishes, mingled with sea-shells. In a deeper layer, 20 to 25 feet below the surface, there occurred the remains of land animals, together with sea-shells of great variety. Croom thought that some teeth belong to the hyena, and Foster reported the hippopotamus; but in both cases the identifications were wrong.

Conrad (Amer. Jour. Sci., vol. XXVIII, 1835, pp. 107–110; Proc. Nat. Inst. Prom. Sci., vol I, pp. 191–192) reported that the bones of animals found here were water-worn, black, and silicified. He concluded that they had been brought down the Neuse River and mingled with sea-shells. The fossiliferous stratum did not rise anywhere more than 10 feet above the river. In the first publication quoted, Conrad published a list of 66 mollusks in this stratum, of which 7 were not yet known as living species and 2 others are noted as new. According to this list, less than 90 per cent are recent. He referred the deposits to his newer Pliocene. In the second publication cited he concluded that the stratum belonged to the post-Pliocene. Stephenson (op. cit., p. 289) refers to the investigations made at this locality. It is not improbable that the deposit which furnished these fossils belongs to the earliest Pleistocene stage, the Nebraskan. The same may be said about the coquina rock mentioned by Stephenson which occurs at Old Fort Fisher, in New Hanover County (op. cit., p. 289, plate XXVIII).

On page 115 the writer refers to a lower jaw of a mastodon found by the geologist W. C. Kerr, near Goldsboro, and described by Joseph Leidy. The jaw was reported to have been found in gravel overlying Miocene marl. The writer believes that the mastodon belonged to the species _Mammut progenium_. Goldsboro, on Neuse River, is near the western border of the Sunderland formation, but the Wicomico is prolonged up the river far above Goldsboro. According to Stephenson and Johnson (op. cit., p. 475), Miocene sands and clays are found over a portion of the northern part of the county (Wayne). The geological age of this mastodon depends more on the age of the gravels in which it was found than on the age of the terrace, although the writer is willing to concede an early Pleistocene stage for the terrace.

A mastodon tooth has been found (see p. 117) somewhere in Wilson County. The county is covered mostly by Pleistocene of Sunderland age, but a small part of the western end is occupied by the Coharie; while, according to Stephenson’s map, both the Chowan and the Wicomico follow up Contentnea Creek into Wilson County. The geological age of the mastodon is doubtful.

At Greenville, Pitt County, have been found remains of _Equus complicatus_, perhaps also of another species of horse (see p. 191). While supposed to have been found in Miocene marls, the tooth belonged without doubt to the Pleistocene. Pitt County is occupied by four Pleistocene formations, Pamlico, Chowan, Wicomico, and Sunderland. The probability is that the horse-teeth were found in an early Pleistocene deposit.

As indicated on page 117, remains of _Mammut americanum_ have been found in Pitt County, possibly at Greenville.

As noted on page 117, a tooth of _Mammut americanum_ has been found at or near Tarboro. Nothing more is known about its origin. At this place are found deposits belonging to the Chowan, Wicomico, and Sunderland formations; it is impossible to say from which the tooth was derived.

Emmons (Geol. Surv. North Carolina, 1852, p. 56) reported finding mastodon bones in marl-pits on the farm of Mr. Knight, on the banks of Tar River, in Nash County, 3 miles west of Rocky Mount. The same Pleistocene deposits occur here as at Tarboro. The bones were supposed to have been buried in Miocene marl, and this may have been true. If so, they belonged to some other species of mastodon than _Mammut americanum_.

On page 191 is given an account of the discovery of teeth of _Equus leidyi_ which were washed up on the beach at Plymouth. This town is on Roanoke River, several miles from Albemarle Sound, and on the border between the Pamlico and the Chowan formations. Our determination of the geological age of the teeth must be based on other evidence than that furnished by the discoverers.

Elsewhere in this work is given an account of finding a part of a skull of a walrus at Kitty Hawk. It was probably during the Wisconsin glacial stage that this animal lived along the coast as far south as Charleston.

As to the geological age of the Pamlico formation, the geologists who have contributed to the report of 1912, the volume cited, hold that it belongs to late Pleistocene. The writer believes that the formation was laid down at a much earlier time. The mastodon jaw and the tooth of _Elephas primigenius_ found in the Inland Waterway Canal may have been buried there during the prevalence of the Wisconsin ice epoch; but, on the other hand, this may have happened during an older Pleistocene stage.

It will be observed that the Pamlico becomes very narrow along the southern third of the coast of North Carolina. In South Carolina it may be represented by one of the older Pleistocene deposits recorded by Sloan; in part possibly by the Wando clays or the Sea island sands. In the author’s view, it is pretty certain that the Pleistocene molluscan fauna which had been found in the Clubfoot and Harlow Canal and at the locality below Newbern corresponds to the Wadmalaw in the vicinity of Charleston. It seems to appear at the southeastern corner of the State, at Southport, and again in the northeastern corner in Dismal Swamp. According to Shaler (10th Ann. Rep. U. S. Geol. Surv., pt. I, 1890, p. 315), a collection of mollusks made near the northern border of the swamp was submitted to Dr. W. H. Dall. There were 29 forms, of which 24 are yet existing, 5 extinct. There were, therefore, 17 per cent of extinct forms. Dall regarded the deposits as belonging to the Pliocene; the writer believes that they may be referred to the Nebraskan stage of the Pleistocene.

From a study of mollusks collected later in the Dismal Swamp Canal, Woolman (Proc. Acad. Nat. Sci. Phila., 1898, pp. 414–428) concluded that they belonged to a time not earlier than late Pliocene and possibly as late as the Pleistocene. Darton (U. S. Geol. Surv., Folio 80) referred the deposits to the Pliocene. Stephenson (op. cit., p. 290) states that recent investigations have led to the conclusion that the beds should be referred to the Pleistocene. The parties in such a dispute may compromise by referring the beds to the Nebraskan stage. It seems probable that the Chowan formation belongs to a stage a little later than these mollusk-bearing beds and represents a strip of old coast marsh, inhabited by elephants, mastodons, horses, and various other animals.

In discussing the causes which led to the production of Cape Hatteras, Professor Shaler (Proc. Bost. Soc. Nat. Hist., vol XIV, 1872, p. 117) remarked that the hard shelly limestone which comes to the surface just above high-tide level along the shore of the mainland from Newbern to the mouth of the Roanoke River looks much like the shell-bed found near Charleston, South Carolina.

SOUTH CAROLINA.

To the reader who wishes to know what work has been done on the Pleistocene geology of South Carolina, two papers may be recommended. The first of these, historical in nature, was published in 1890 by Professor Joseph A. Holmes (Jour. Elisha Mitchell Sci. Soc., vol. VII, pp. 89–117), the second in 1905 by Dr. Griffith T. Pugh (Thesis, Vanderbilt Univ., pp. 1–74). Those who have contributed most to a knowledge of the palæontology of this formation are Tuomey, F. S. Holmes, Leidy, Dall, Dall and Harris, Earle Sloan, and G. T. Pugh. J. A. Holmes, Tuomey, F. S. Holmes, and Dall have made important contributions to the knowledge of the invertebrate animals. For our knowledge of the vertebrates we are indebted principally to F. S. Holmes and Joseph Leidy. The author who has dealt most recently and in considerable detail with the stratigraphy of the Pleistocene deposits is Earle Sloan, State geologist (Bull. No. 2, ser. IV, South Carolina Geol. Surv., 1908, 479 pages). From these authorities we learn that, while the larger part of the Coastal Plain may be to a greater or less extent overlain by deposits referable to McGee’s Columbian, the deposits which bear fossils are confined almost wholly to a narrow strip along the coast. In this strip have been found the numerous mollusks listed and described by Tuomey, F. S. Holmes, and W. H. Dall, as well as most of the species of vertebrate fossils. The fossiliferous deposits do not usually extend back from the coast more than about 10 miles.

Undoubtedly fossil-bearing Pleistocene deposits are to be found here and there along all the rivers, perhaps to the western border of the Coastal Plain. This is indicated by the discovery of remains of horses and mastodons in Darlington and Richland counties. The thickness of the Pleistocene deposits along the coast is said to amount to as much as 60 feet, but it is usually much less. Only a part of this is fossiliferous, a bed that appears to vary in thickness from about 3 to 8 feet. This is found as much as 8 feet above mean-tide level, sometimes below it. The materials of this fossiliferous bed vary greatly. Sometimes they consist almost entirely of shells of mollusks, in other cases of a blue mud or sand, and with these may be mingled peaty materials, gravel, and again rolled masses derived from the underlying deposits. The fossils contained in the bed mentioned consist of mollusks, and in some places bones and teeth of vertebrates occur in more or less abundance. The bed is underlain often by deposits of Tertiary age. Bones and teeth of the vertebrates, as fishes and cetaceans, that lived when those Tertiary rocks were being deposited may occasionally have been washed into the Pleistocene bed. Again, where the older and the newer beds are exposed along the shores, fossils may be washed out of both and commingled on the beach; then again, a great part of the fossils collected along this coast of South Carolina have been rescued from the phosphate rock gathered for commercial purposes. This has been to a great extent dredged from the rivers; and thus remains of Pleistocene and of Tertiary animals have been mixed indiscriminately together. It is often impossible to determine to what formation a fossil may belong. To add to the difficulty of the palæontologist, the vertebrate remains are sometimes found washed out and mingled with bones or teeth of what appear to have been domestic animals.

Beginning at the northern end of the South Carolina coast-line, the first locality furnishing Pleistocene fossils is, or rather was (Pugh, op. cit. p. 33), White (or Price’s) Creek, in Horry County. Here at a height of about 5 feet above tide was found a bed approximately 6 feet thick apparently thrown up on the shore by storms (Tuomey, Geol. Rep., 1848, p. 187). No vertebrates have been reported from the locality. At Laurel Hill, in the extreme northeastern corner of Georgetown County, Tuomey (op. cit., pp. 187, 188) found a perpendicular bluff 30 feet high, at the base of which was a bed 8 feet thick made up of sand and broken shells. The top of the bed was 8 feet above tide, the highest elevation reached by the bed along the South Carolina coast. Tuomey mentions other localities around Georgetown where the fossiliferous bed was discovered. One was on Santee River. No vertebrates appear to have been met with in this region. In Christ Church parish, in Charleston County, Tuomey discovered several exposures of the bed in question, and this was sometimes so superficial as to be within reach of the plow.

Pugh (Pleistocene Deposits, etc., p. 34) quotes from F. S. Holmes a section which was found at Goose Creek, north of Charleston, as follows:

Yellow sand 12 feet Blue mud 29 feet Ferruginous sand, containing bones, etc. 3 inches Yellow sand 3 feet Pliocene marl resting on Eocene white marl 12 feet

The bones occurred likewise in the blue mud, and such were especially well preserved. Leidy (Holmes’s Post-Pl. Foss. S. C., p. 102) recounts his observations at this locality; nevertheless, the only vertebrate fossil that the writer finds credited by Leidy to this locality is a tooth of _Equus fraternus_ (=_E. leidyi_), which he figured (plate XV, fig. 8).

Dredging for phosphate rock has been carried on extensively in Cooper River; but of Pleistocene vertebrate fossils secured here the writer has record of only _Megatherium, mirabile_. This is represented in the Charleston Museum by a portion of a lower jaw.

Wando River is situated northeast of Charleston, runs parallel with the coast, and empties into Cooper River. From this have (according to the writer’s knowledge) been secured only _Equus complicatus_ and a part of a tusk of _Odobenus_. The latter is in the Charleston Museum. In most cases no record has been kept of the origin of the specimens in collections.

The Pleistocene bed along Ashley River is famous for the number of fossil vertebrates which it has furnished. It has been described by F. S. Holmes in various publications, especially in the Introduction to his Post-Pleiocene Fossils of South Carolina, 1860, pages I-XII. In the same work, on pages 99–100, Dr. Leidy briefly described the geological character of the beds; and on subsequent pages he described the vertebrate species found there. The principal beds were located on Ashley River, about 10 miles above Charleston. According to Pugh (“Pleistocene Deposits of South Carolina,” p. 34), the fossiliferous deposits rest on Miocene marls. At the top are 4 feet of yellow sands with bands of clay; below, is a foot or more of blue mud lying on the Miocene. The bones are more numerous and best preserved in the blue mud. The Pleistocene bed is elevated only a few feet above tide-level. Inasmuch as nearly all the species of Pleistocene vertebrates which have been found along the South Carolina coast have been secured along the Ashley River, the few found elsewhere will be included in the following list. Some of those marked found somewhere about Charleston may have been collected in or along Ashley River. In this list the contractions following the names signify as follows: A, Ashley River; B, the region about Beaufort; C, somewhere around Charleston; C. r., Cooper River; E, Edisto River; G. c., Goose Creek; J. i., John’s Island; S. r., Stone River; W. r., Wando River; Y., Yonge’s or Young Island. The species preceded by the dagger are extinct.

Odobenus rosmarus A., W. r. (p. 29). Lynx ruff us C. †Canis sp. indet. C. Procyon lotor A. †Arctodus pristinus A. Ursus americanus C. Sylvilagus floridanus? A. †Hydrochœrus æsopi A. †Hydrochœrus pinckneyi C. (p. 365). †Castoroides ohioensis A. (p. 279). Castor canadensis A. Ondatra zibethica A. †Elephas imperator C. (p. 162). †Elephas columbi A., B. (p. 155). †Mammut americanum A., B. (p. 118). Mammut progenium (p. 118). †Bison latifrons? A. (p. 260). †Bison sp. indet. A. (p. 260). †Alces runnymedensis C. (p. 364). Cervus canadensis A. (p. 242). Odocoileus virginianus? A. (p. 231). Camelops sp. indet. †Tagassu lenis A. (p. 222). †Tagassu sp. indet.? A. (p. 222). †Tapirus haysii A. (p. 204). †Tapirus sp. indet. A. (p. 205). †Equus complicatus A., W. r., B. (p. 192). †Equus leidyi A., J. i., G. c., S. r., B. (p. 192). †Equus littoralis C. (p. 193). †Hipparion venustum A. †Physeter vetus A. †Trichechus antiquus A. †Megatherium mirabile A., C. r., S. r. (p. 35). †Mylodon harlani A. (p. 35). †Megalonyx jeffersonii B. (p. 35). †Didelphis virginiana J. i. †Alligator mississippiensis A. †Pseudemys sp. indet. A. †Testudo crassiscutata? A. Trichiurus lepturus Y. †Istiophorus robustus Y. †Ischyrhiza mira? A. Lepisosteus osseus A. Dasyatis hastata? Y.

Besides the species enumerated, the early collectors found remains which were identified as belonging to such domestic animals as the dog, ox, sheep, and hog. Leidy rejected these as Pleistocene species, while Holmes and Agassiz accepted them as such. Possibly the supposed dog was in reality a wolf and the supposed ox a bison. Small teeth like those of cows are fossilized as are the teeth of extinct animals. At Bee’s Ferry on Ashley River the fossiliferous bed has a thickness of 3.5 feet and is at about high-water mark. It is overlain by from 15 to 20 feet of loose sands.

By far the most of the species have been entered in the list on the authority of Joseph Leidy. Only F. S. Holmes reported the elk (_Cervus canadensis_), and the writer has seen two teeth of the species at the Academy of Natural Sciences in Philadelphia credited to Charleston. Holmes also reported _Glyptodon_, but that is not included in the list. _Lynx ruffus_, _Ursus americanus_, _Hydrochœrus pinckneyi_, _Elephas imperator_, _Bison latifrons_, _Alces runnymedensis_, _Camelops_ sp., and _Equus littoralis_ are included on the evidence of specimens seen by the writer in the Charleston Museum or in some of the other collections made on the coast of South Carolina. Loomis has recently (Amer. Jour. Sci., vol. XLV, 1918, p. 438) described a specimen of _Mammut progenium_ (as _Mastodon americanus_) from near Charleston and another from near Beaufort.

_Alces runnymedensis_ was first briefly referred to in Year Book No. 14 of the Carnegie Institution of Washington, 1915 (1916), page 387. The name is based on an upper right hindermost milk molar in the Charleston Museum (No. 13534). It is the property of Mr. Charles C. Pinckney. Where the tooth was found is not known, but it was somewhere near Charleston, in the phosphate-bearing area. The specific name is that of the estate of the owner. The tooth closely resembles the corresponding one of _Alces americanus_, but is larger and has a flatter crown. Only the crown of the tooth is preserved, and of this a part of the enamel of the inner anterior cone is broken off; otherwise it is in fine condition. The color is very black. The following measurements are given of this tooth and of the corresponding one of _Alces americanus_, No. 117055 of the U. S. Biological Survey. The two teeth are only slightly worn.

_Measurements of milk molars of Alces, in millimeters._

┌────────────────────────────────────┬──────────────┬─────────────────┐ │ Dimensions taken. │A. americanum.│A. runnymedensis.│ ├────────────────────────────────────┼──────────────┼─────────────────┤ │Length of tooth near outer border │ 24.0│ 25.5│ │Length of tooth at middle width │ 21.5│ 23.0│ │Width of tooth along front border │ 23.0│ 23.0│ │Width of tooth from median style to │ │ │ │ base of inner hinder cone │ 21.0│ 24.0│ └────────────────────────────────────┴──────────────┴─────────────────┘

The angle between the outer and inner faces of the hinder half of the tooth is 54° in the tooth of the existing species, 64° in the fossil tooth. On the grinding-surface the fossettes are wider than in the tooth of the existing moose.

It is interesting to find this moose in the region about Charleston. We must suppose that it lived there during one of the glacial stages, probably when the walrus occupied that part of the coast.

In the Pinckney collection is a tooth of a capybara that deserves attention. A figure of it is here presented (fig. 18), a side view. Exactly where the tooth was found is not known, but it was somewhere in the vicinity of Charleston. The tooth is the upper left hindermost molar. In the figure the front end is directed toward the left hand. There are present 17 plates. None of the plates either in front or behind are missing. The free edges of the plates are not turned backward. The length of the tooth is 62 mm., the width is 17.5, the height of the plates on the inner face 37 mm., but probably the less calcified bases of the plates have been destroyed.

On the grinding-surface the plates run obliquely from the inside outward and backward. As seen on the inner face, the plates, as they pass to the grinding-surface, lean backward. The corresponding tooth of a capybara from Surinam has a length of 37 mm. The length of its skull from foramen magnum to the front of the snout is 215 mm. In case the skull of the fossil was long in proportion to the length of the tooth, the length as given above would be 360 mm., about 15 inches.

To this fine large species I give the name _Hydrochœrus pinckneyi_, in honor of Mr. Charles C. Pinckney, the owner of a collection of fossils from the region about Charleston and the proprietor of the estate of Runnymede, near Lambs, South Carolina.

In the same collection is a part of the lower jaw, right side, of a rather large wolf. In this jaw there remain the complete fourth premolar, the roots of the third premolar, and one root of the second (fig. 19).

The following measurements are taken from the fragment mentioned; from the corresponding part of a jaw of _Ænocyon dirus_, No. 8307, from La Brea, California; from the gray wolf, _Canis occidentalis_, from Fort Simpson, British America, No. 9001, U. S. National Museum; and from the type of _C. floridanus_, in the U. S. National Museum.

_Measurements of jaws and teeth of wolves, in millimeters._

┌───────────────────────┬──────────┬──────────┬────────────┬──────────┐ │ Parts measured. │ │ │ C. │ C. │ │ │Charleston│ La Brea │occidentalis│floridanus│ │ │ jaw. │ jaw. │ jaw. │ type. │ ├───────────────────────┼──────────┼──────────┼────────────┼──────────┤ │Height of jaw in front │ │ │ │ │ │ of pm_{4} │ 28│ 32│ 33│ 21.5│ │Thickness at front of │ │ │ │ │ │ pm_{4} │ 14│ 16│ 14.2│ 10.2│ │Length of pm_{4} │ 18.5│ 20.2│ 18.5│ 14.5│ │Thickness of hinder │ │ │ │ │ │ lobe of pm_{4} │ 9.5│ 11│ 9.5│ 7│ │Thickness of front lobe│ 8.5│ 9.8│ 8.5│ 6.4│ └───────────────────────┴──────────┴──────────┴────────────┴──────────┘

The measurements show that the fossil is much too large to belong to the wolf now inhabiting Florida. It appears also to be too small to belong to the wolf _Ænocyon dirus_, and _A. ayersi_ was but little if any smaller. The lower teeth of the latter species are not known. The accordance in measurements with those of _C. occidentalis_ makes it probable that the fossil jaw found at Charleston belonged to a wolf not greatly different. With the materials at hand it is impossible to refer the jaw specifically.

Within the city of Charleston the bed bearing vertebrate fossils is said to be several feet below tide-level. At Young Island, Wadmalaw Sound, nearly 20 miles southwest of Charleston, the top of the fossil-bearing stratum is at tide-level. This locality is otherwise known in the literature as Simmons’s. The only Pleistocene vertebrate fossils that the writer finds reported from the place are the fishes _Lepisosteus osseus_ and _Trichiurus lepturus_.

In the region about Beaufort, the same fossil-bearing stratum, having about the same composition and the same elevation, is met with in many places. A few species of fossil vertebrates and many invertebrates have been secured. Here have been found _Mammut americanum_ (p. 118), _Elephas columbi_ (p. 155), _Equus complicatus_ (p. 191), and _Megalonyx jeffersonii_ (p. 35).

A brief notice will be taken of the few known localities where, away from the immediate coast, vertebrate fossils have come to light.

Tuomey, in 1848 (Rep. Geol. South Carolina, p. 177), in describing marls found near Darlington, on the farm of G. W. Dargan, and which he regarded as belonging to the Pliocene, reported the discovery of two perfect molars of a mastodon (p. 118). The locality was in a swamp, and the bed of marl was covered with 3 or 4 feet of black mud. The teeth were immediately below the mud and enveloped in the marl. These teeth belonged to _Mammut americanum_ and had been deposited at some time during the Pleistocene. At another place fragments of the antlers of a deer were found in the marl. In such cases the marls formed at one time the surface of the ground, or more probably the bottom of a swamp; and the Pleistocene bones and teeth might have been trampled down into the marl by living animals. On page 119 is given an account of another mastodon tooth discovered in the same county; and the teeth of a horse have been reported as having been found, associated with those of the mastodon (see p. 193).

In Lee County, adjoining Darlington County on the southwest, at a locality “near Concord church,” between Lynch’s Creek and Black River, Tuomey (op. cit., p. 178) found a bed of Pliocene marl about 4 feet thick. From an excavation in this marl had been taken a tusk which Tuomey regarded as that of a mastodon, but this may have belonged to an elephant. In Berkeley County, at the head of Cooper River, there is, or was, a morass known as Biggin Swamp. This was passed through in constructing the Santee Canal. On page 156 is an account of the discovery of remains of _Elephas columbi_ and of _Mammut americanum_; on page 162, the finding of a tooth of _Elephas imperator_. The discovery of the latter marks the age of the deposits as being about that of the Aftonian interglacial.

It has been seen that at many points along the coast there is a fossiliferous stratum varying from 2 to 8 feet. At most localities the fossils consist principally of marine animals, especially mollusks, and the deposits have evidently been laid down in salt water. Along Ashley River and at some localities in the region about Beaufort it seems evident that the surface was above, but not far above, sea-level, and that it formed a swamp on which a great variety of land animals could move about and feed. After death their bones would suffer the fate which befalls them in such cases. Most of them would undergo decay. Parts would be trampled into the muck, broken into fragments, and undergo still further decay. Only the most durable parts, as the teeth, antlers, and the more solid bones would usually stand a chance for preservation. Apparently, on this coast, no considerable parts of one skeleton have ever been found, or at least reported. In Charleston Museum are many bones of a skeleton of _Megatherium_, but it is uncertain where it was found.

The list of vertebrates referred to the Pleistocene of the South Carolina coast contains 33 species of mammals, of which 24 appear to be extinct. This high proportion of extinct species seems to confirm our reference of the fauna to the early Pleistocene. Besides the extinct forms, it is to be noted that within historical times the muskrat, beaver, and elk have not lived in the region about Charleston.

Pugh (Pleist. Deposits S. C., p. 66), from a study of the Pleistocene marine mollusca of South Carolina, has concluded that, if the Pleistocene sea-temperature differed at all from that of the present, it was slightly higher rather than slightly lower. It must be remembered, however, that the Pleistocene represented a very long period of time and that, farther north, the climate underwent great fluctuations. That these fluctuations would not have affected the temperature of the sea along the coast of the Carolinas is not probable. It is hardly supposable that capybaras and manatees lived about Charleston at the same time that the moose and the walrus were there. The latter had been forced down there during some glacial stage, possibly the Wisconsin; while the horses, tapirs, elephants, manatees, the mylodon, and the megatherium had their existence, we may suppose, about the time of the Aftonian. During this stage, too, lived the species of mollusks which Pugh has elaborated. It would seem that after that time some change took place in conditions, probably a slight elevation, so that little more than beds of unfossiliferous sand and marls were deposited.

Professor Earle Sloan, in his “Mineral Localities of South Carolina” (Bull. No. 2, ser. IV., South Carolina Geol. Surv.), has recognized the following divisions in the marine Pleistocene of the State:

6. Sea Island loams.

5. Wando clays and sands.

4. Accabee gravels.

3. Bohicket marl-sands.

2. Wadmalaw marl.

1. Ten-Mile sands.

Of these, the fossiliferous deposits referred to above appear to belong to the Wadmalaw marl. It may be confidently expected that somewhere along the South Carolina coast, beneath the beds bearing the vertebrate fossils, there will yet be discovered other Pleistocene deposits, probably shell marls, which belong to the Nebraska stage.

GEORGIA.

The only part of Georgia at present of interest to the student of vertebrate palæontology is that which lies immediately along the Atlantic coast and along a few of the larger rivers. The northwestern corner of the State is mountainous and probably contains little or no Pleistocene. The Coastal Plain extends landward to a line which starts at Augusta, on Savannah River, passes through Milledgeville and Macon, and ends at Columbus, on the Chattahoochee. A large part of this region is mantled by a deposit resulting from the decay of the underlying rocks. These deposits are of uncertain age, a part belonging probably to the Pleistocene, but the large part to the Pliocene or to still older Tertiary. The Pleistocene has not yet been differentiated from the remainder, and, in any case, has furnished no vertebrate fossils. For information on the subject the reader may consult McGee (12th Ann. Rep. U. S. Geol., Surv., pt. I, pp. 478–484), Spencer (Geol. Surv. Georgia, 1890–91, pp. 61–81), and Veatch and Stephenson (Bull. 26, Geol. Surv., Georgia, pp. 400–456).

The deposits in Georgia which can with certainty be referred to the Pleistocene form a broad belt lying along the coast and extending landward a distance of about 30 miles along Savannah River and about 60 miles at the Florida boundary line. For a description of these deposits the reader is referred to Veatch and Stephenson’s article in Bulletin 26 just mentioned, pages 424–456. These deposits are disposed in two terraces, a higher and older and a lower and younger. The older is named the Okefenokee formation, the younger the Satilla formation. The positions of these may be observed in the figure here presented, taken from Bulletin 26 above referred to (fig. 20).

The Okefenokee terrace has a breadth of 20 to 40 miles and an elevation of 60 to about 125 feet above sea-level. It forms a plain which Veatch and Stephenson describe as in general flat and almost featureless. It is dotted with cypress ponds and swamps, with here and there low ridges and hills of sand. Along the larger streams which cross the plain are found terraces supposed to have been laid down while the Okefenokee terrace was forming; they extend far back into the State. In neither the main terrace nor the fluviatile terraces have any fossils been found, except a little silicified wood.

The Satilla Plain extends backward from the coast 20 to 30 miles and varies in elevation from 15 to 40 feet. On the landward side it ends in an escarpment which is taken, by the authors quoted, to be an old sea-beach. Along the large rivers it is continued as a series of terraces occupying a lower position than those of Okefenokee time. According to Veatch and Stephenson, this formation consists of unconsolidated clays, sands, and thin layers of gravel. The thickness averages about 15 feet, but may become as much as 45 or 50 feet.

The Satilla deposits are fossiliferous. At various places, at some distance from the coast, sea-shells occur, especially shells of oysters. This shows that at times the plain, or at least some parts of it, has been under sea-water. Bones and teeth of vertebrate animals have been discovered at several localities, but at only two places have identifiable materials been secured. The region about Brunswick and that just south of Savannah have furnished important collections of vertebrate animals.

During the years 1838 and 1839 an attempt was made to construct a canal to connect Altamaha River with Turtle River at Brunswick. Some bones of large mammals were met with and came to the notice of Hamilton Couper, and through him became known to the scientific world. The most striking was the great ground-sloth, of the genus _Megatherium_, and which Leidy afterwards called _Megatherium mirabile_. At a more recent time, during dredging operations, probably in the harbor, other remains were found and turned over to the Geological Survey of Georgia. The fragmentary bones and teeth were identified by Mr. J. W. Gidley (Bull. No. 26, Geol. Surv. Georgia, p. 436).

The fragments of teeth regarded by Gidley as belonging to _Mammut floridanum_ appear to the writer to represent _Gomphotherium rugosidens_, a species rather common in that region and belonging to the upper Miocene or Lower Pliocene. Four teeth identified as those of _Physeter vetus_ or _Physeterula neolassicus_ appear to be identical with Leidy’s _Orycterocetus quadratidens_; but this may be possibly the same as _Physeterula neolassicus_ (=_P. dubusi_). It, too, is older than the Pleistocene. From the two collections have been determined the following list:

Castoroides ohioensis (p. 280). Elephas columbi (p. 157). Mammut americanum (p. 120). Bison sp. indet. (p. 261). Cervus? sp. indet. (p. 243). Tapirus haysii (p. 206). Equus complicatus (p. 193). E. leidyi (p. 193). E. littoralis (p. 193). Megatherium mirabile (p. 36). Mylodon harlani (p. 36). Chelonia (??) couperi. Crocodylus (?) sp. indet. Lamna sp. indet. Galeocerdo sp. indet. Carcharodon sp. indet. Dasyatis sp. indet.

With the bones found in the canal was a femur 13 inches long, which Harlan described as _Chelonia couperi_, but which resembles more closely that of some edentate mammal. Gidley stated that the shark-teeth probably represent Eocene and Miocene species. This may be true, but the supposition is not necessary, inasmuch as species of all three genera are yet living on our Atlantic coast.

J. Hamilton Couper (Hodgson’s Memoir, pp. 37–40) has given an account of the topography and geology of the region through which the Brunswick Canal was being constructed (map 40). On one of the plates of the work is a section from the ocean westward 21 miles. About 10 miles west of St. Simon’s Island the canal passed through Six-mile Swamp. This is connected at its northern end with Altamaha River, at the southern with Turtle River. The swamp has thus the appearance of a lake which has become filled with alluvial deposits. These consist of a compact clay, usually yellow and impregnated with iron. There are thin strata of soft, chalky marl and many fragments of petrified wood. At the bottom of this deposit were found the bones of _Megatherium_, _Elephas_, _Mammut_, _Equus_, and _Bison_. Beneath the clay stratum was sand with marine shells. Overlying the clay was a thin stratum of vegetable and sandy loam. The bones occurred at a depth of from 4 to 6 feet. In no instance, except when they had been washed out into the salt-water creek, was there any abrasion of the surface or incrustation of marine shells.

The geologist Charles Lyell (Second Visit, etc., vol. I, p. 347) stated that part of a skeleton of a megatherium, dug out in cutting the canal, was so near the surface that it was penetrated by the roots of a pine tree. As a considerable number of the bones of one skeleton were found together, Lyell supposed that a whole carcass had been floated down the river to the spot.

Even before remains of fossil vertebrates had been found at Brunswick, bones had been discovered at Skidaway Island, near Savannah. As early as 1823, S. L. Mitchill (Ann. N. Y. Lyc. Nat. Hist., vol. I, p. 58) announced the finding of teeth of _Megatherium_ at this place. More than 20 bones of the same animal were reported from the same locality in 1824 by William Couper. In 1846 (Hodgson’s “Memoir on Megatherium,” pp. 25–30), Dr. Joseph Habersham published a list of the species discovered up to that time. Lyell (Second Visit, etc., vol. I, p. 313) gave an account of his visit to the locality and noted the species obtained. The following list appears to contain all found there:

Elephas columbi (p. 157). Mammut americanum (p. 120). Bison sp. indet. (p. 262). Equus leidyi? (p. 194). Megatherium mirabile (p. 36). Mylodon harlani (p. 37). Terrapene canaliculata.

The box-tortoise _Terrapene canaliculata_ was described by the writer in 1907 (Bull. Amer. Mus. Nat. Hist., vol. XXIII, p. 850, figs. 5–7) on fragmentary materials found in the U. S. National Museum. These had been sent there by Dr. J. P. Scriven, who had been active in collecting the fossil vertebrates about Savannah. Whether the remains of this box-tortoise were found on Skidaway Island or in Whitemarsh Island is uncertain.

Besides these species, found on Skidaway island, two species, _Mammut americanum_ and _Mylodon harlani_, have been found at Heyner’s (or Hainer’s) Bridge. This is about 7 miles south of Savannah, where the road crosses Vernon Creek (Lyell, “Travels in North America,” vol I, pp. 163–164). Here the stream is called White Bluff Creek. In order that the reader may get a clear understanding of the conditions at this important locality, a map found in Hodgson’s Memoir is reproduced (map 40).

The whole region south of Savannah, between the mouths of Savannah and Ogeechee Rivers, is low and much divided into islands by streams connected with the rivers mentioned. A considerable part of these islands consists of marshes, which are usually overflowed by the tides. Most of the fossil bones were found along the southern bank of Skidaway River, in two places, apparently about 0.5 mile apart and near the western end of the island. On the map Hodgson has named the locality Fossilossa. Here Skidaway River made a bend which caused the bank to be eroded away, thus exposing the bones. According to Couper (Hodgson’s Memoir, p. 40), the bones were embedded in the marsh formation at about the level of very low-water. Lyell (Second Visit, etc., vol. I, p. 314) stated that the bones occurred in a dark peaty soil, or marsh mud, above which was a stratum of sand 3 or 4 feet thick; while below the peaty soil and below sea-level was sand containing many marine fossil shells, all belonging to species yet living on the neighboring coast.

The authors quoted state that at various places along the Georgia coast are found stumps of trees, cypress, cedar, and pine, in the deposits of the salt marshes and at a depth of from 2 to 4 feet below high-water. This is taken as evidence of subsidence in that region.

It is a matter of importance to know how those animal remains reached their place of burial. It has been suggested that whole carcasses had been floated down the streams and sunken where the bones are found. This is possible, but not probable. The peaty nature of the deposit inclosing the bones appears to be opposed to this view; nor could disarticulated bones have been washed down far from above, for they show no signs of attrition. The most probable explanation is that these animals lived and died about where their bones were discovered. At some past time the surface stood at a higher level than at present, although low enough to be more or less marshy. It probably supported a dense forest growth, and hither the species listed above resorted, with many others not yet discovered.

The animals inhabiting the region represent the same fauna found at so many places in Florida and Texas. The writer believes that they existed during the early part of the Pleistocene, approximately during the Aftonian interglacial; and that some of the species, as _Megatherium_, _Mylodon_, _Equus_, and _Tapirus haysii_ became extinct before the advent of the Wisconsin glacial stage, probably a long time before this.

FLORIDA.

(Maps 7, 8, 15.)

For the most recent descriptions of the geology of Florida one must consult the Annual Reports of the Florida Geological Survey, issued by the State geologist, Dr. E. H. Sellards, and Water supply Paper 319 of the U. S. Geological Survey, prepared by George C. Matson and Samuel Sanford and published in 1913. In the latter work are two large maps, one representing the topography of the State and the distribution of the various geologic formations; the other presents a generalized view of the distribution of Pleistocene terraces, as recognized by Matson and Sanford. The Second Annual Report of the Florida Geological Survey contains a map similar to the first mentioned.

From these maps it will be seen that the surface of Florida is largely occupied by Pleistocene deposits. According to Matson and Sanford, these deposits present themselves as disposed mostly in three principal terraces; and these are believed to indicate that the State was at one time largely submerged beneath the sea and that its present condition was attained after three principal upward movements. As shown on plate V of the geologists just named, the northern half of the peninsula at the time of greatest depression was represented by a number of islands, two of considerable size. One of these was situated at the northern end of the peninsula, the other near its center. The materials laid down around these islands and bordering the dry land along the northern border of the western half of the State form what is called the Newberry terrace. Its surface stands now at a height varying from 70 to somewhat more than 100 feet above sea-level. A second elevation exposed the deposits which, at least in part, constitute the next terrace, the Tsala Apopka. Its surface is a plain having an elevation of 40 to 60 feet above sea-level. At this stage the islands of the peninsula had coalesced, and the dry land extended southward nearly to the present Lake Okeechobee. A broad belt along the Atlantic and Gulf coasts, now dry land, was still occupied by salt water. A third elevation of the land left exposed the lowest terrace, the Pensacola, that bordering the Atlantic and Gulf coasts and including the southern end of the peninsula somewhat farther north than Lake Okeechobee. The elevation of this terrace varies from that of sea-level up to about 40 feet.

The materials composing the terraces in Florida are principally sand with local deposits of clays. In the southern part of the State important beds of limestone are found in the Pensacola terrace. These beds are shown on Matson and Stanford’s geological map. At St. Augustine and along the coast southward are beds of sea-shells cemented into coquina. Where cementation has not occurred there are beds of loose shells and of marl and sand.

The writer has already (p. 346) expressed his opinion regarding the Coastal Plain terraces found in the States farther north. He finds in Florida nothing to contradict, but much to confirm, that opinion. Whatever may be the origin of Newberry and Tsala Apopka terraces, they were not laid down in salt water. From the descriptions of the deposits there the stratification and the alternation of the materials do not exist that one might expect; but, above all, there seem to be no marine fossils to attest to the presence of the sea. In Florida, too, here and there over these higher lands there are found, in place of marine fossils, the remains of many extinct land animals, as mastodons, elephants, horses, ground-sloths, and the like.

As regards the Pensacola terrace, there are found at its base, within a few feet above or below sea-level, deposits containing remains of such animals as have just been mentioned, besides many others. Often the state of preservation of these remains and the condition of their burial are such that we must conclude that the animals lived and died on the spot. Furthermore, these animals constitute an assemblage corresponding to that found in western Iowa, in Nebraska, and in Oregon, which are believed to have existed during the first interglacial stage. It corresponds also to that met with under similar conditions and at the same level at Savannah, at Charleston, at Brunswick, and at Long Branch. In most cases, too, this fossiliferous stratum is overlain with very scant deposits. By some geologists and palæontologists the animals are regarded as belonging to the Pliocene.

If the reference of the fossil vertebrates mentioned is not wholly wrong, it follows that the lowest terrace or plain along the coast was not laid down late in the Pleistocene, but at an early stage, and the higher plains must have been formed at still earlier times.

At Vero, as will be shown on page 382, a large assemblage of fossil vertebrates has been secured. The bed furnishing the oldest fossils, those of the bed known as No. 2 and believed to be of about first interglacial age, is underlain by a bed of marine shells, also of Pleistocene age. This bed is regarded by Dr. E. H. Sellards as being equivalent to the coquina which is so well known at St. Augustine; and the same formation is found here and there along both coasts of the peninsula (Matson and Sanford, op. cit., p. 192). Probably not all deposits that are called coquina are of the same age, but the deposits in question pass, on the landward side, beneath the deposits which bear vertebrate fossils. The bed at Vero, No. 2, must have been laid down after an uplift had brought above sea-level the bed of shells No. 1, on which No. 2 reposes; that is, between the time of deposition of No. 1 and No. 2 there must have elapsed a considerable interval of time. The shell deposit, therefore, probably belongs to the first glacial epoch, the Nebraskan. Inasmuch as a similar vertebrate fauna is found on both the eastern and the western coasts of the peninsula, it follows that any Pleistocene deposits underlying these vertebrate-bearing beds belongs to the Nebraskan stage; in places these have great thickness. Matson and Sanford (op. cit., pp. 194–195) concluded that the maximum thickness of the Pleistocene in southern Florida, disregarding the sandhills, is probably about 125 feet. Even if it were a matter of importance to determine in or on which terraces the vertebrate fossils are found, it would not always be easy to do so. The majority of specimens have been discovered around the coasts of the State, and therefore in deposits referred to the youngest terrace. In other cases it is difficult to determine the terrace in which fossils are buried, partly because of imperfect records as regards locality, kind of deposits, and depth of burial, partly because each terrace extends up the river valleys beyond its general border. The various fossil-bearing localities will therefore be taken up by counties, beginning at the western end of the State and ending at the southern end.

_Jackson County._—As already recorded on page 121, a tooth of _Mammut americanum_ has been found at Marianna. No details have been recorded. The Newberry terrace extends nearly or quite to this town. If it could be shown that this tooth had been buried in that terrace when it was formed, it would probably have to be referred to the time of the first glacial stage.

_Gadsden County._—It appears that no vertebrate remains belonging to the Pleistocene have been found in this county, except a tooth of _Mammut americanum_ (p. 157) which was discovered somewhere in Little River.

_Wakulla County._—On page 157 the finding of a tooth of _Elephas columbi_ somewhere along St. Marks River has been mentioned; also the discovery of a part of a skeleton of either a mastodon or an elephant somewhere about Wakulla Springs.

_Columbia County._—A mastodon tooth has been found in this county 3 miles northwest of Fort White (p. 121). To which terrace it belonged or what is its place in Pleistocene time it is impossible to say.

_Nassau County._—At Stokes Ferry have been found some teeth of an extinct horse (p. 194), a fragment of a tooth of an elephant (p. 180) and some ear-bones of a whale. Veatch and Stephenson (Bull. 26, Geol. Surv. Georgia, p. 394) report that these appeared to come from either the Charlton formation or the Satilla. If the Charlton really belongs to the Pliocene it is not probable that the fossils were derived from it; if they were derived from the Satilla, they do not belong to late Pleistocene.

_Duval County._—On page 106 of the Eighth Annual Report of the Florida Geological Survey, Sellards reported the finding of remains of _Mammut americanum_ (p. 122), _Elephas columbi_ (p. 157), an undetermined species of _Bison_ (p. 262), and an undetermined species of _Odocoileus_ (p. 232), near Pablo Beach, at station 120 on the Inland Waterway Canal. Here, too, has been discovered a bone of _Trachemys? nuchocarinata_. Sellards stated that the position of the beds here is the same as that of the other localities along the Atlantic coast, the fossils being found in sand and muck which rest upon Pleistocene shell-marl. The locality is, of course, on the youngest terrace; but that, in the opinion of the writer, belongs to the early Pleistocene.

_St. John’s County._—At a place 28 miles south of St. Augustine, along the Inland Waterway Canal, Mr. Fred P. Allen, of St. Augustine, collected on the Almero farm remains of _Mammut americanum_ (p. 122), _Elephas columbi_ (p. 158), _Mylodon harlani?_ (p. 37), _Equus_ sp. indet. (p. 194), the box-tortoise _Terrapene antipex_, and a dermal plate of perhaps _Alligator mississippiensis_. These were found in the banks of the canal. Here, at least, the horse and the mylodon, taking into consideration the geological circumstances, indicate early Pleistocene, equivalent to the first interglacial stage.

_Levy and Alachua Counties._—Geologically these counties furnish important localities because of the presence of the Alachua clays (usually referred to the lower Pliocene or even the Upper Miocene) and deposits belonging to all three of the Pleistocene terraces, Newberry, Tsala Apopka, and Pensacola. The Alachua clays first require consideration, for in them have been found a considerable number of species of vertebrates which usually indicate Pleistocene deposits. The localities where Alachua clays have furnished vertebrate fossils, as indicated on Matson and Sanford’s map (Water Supply Paper 319, U. S. Geol. Surv., plate I), are situated, one around Archer, Alachua County (the type locality), second, about 5 miles west of Williston, in Levy County, and a third about 5 miles east of Newberry, in Alachua County.

The clays referred to form accumulations in depressions on the surface of the Ocala limestone, itself belonging to the Eocene. The deposits are said to average in depth about 10 feet, but are often thinner and occasionally much thicker. They have furnished a considerable number of species of vertebrates. A list, prepared by Dr. Leidy, of those found at Archer was published in 1892, in Bulletin 84 of the U. S. Geological Survey, on page 129. Besides these, Leidy had previously reported a tapir, a small crocodile or alligator, and a bone thought to belong to the extinct _Cervus americanus_ (_Cervalces scotti?_), but which was not afterward mentioned. The rhinoceroses and the camels were described by Leidy and Lucas in 1896 (Trans. Wagner Free Inst., vol. IV, pp. 1–61 with plates).

Herewith is presented a list of such vertebrates as have been found at Archer. It appears necessary to retain for the rhinoceroses the specific names given them by Leidy.

Gomphotherium floridanum (p. 121). Odocoileus osceola? (p. 232). Procamelus major (p. 224). P. minor (p. 224). P. minimus (p. 224). Teleoceras proterus (p. 211). Aphelops longipes (p. 211). Tapirus haysii? (p. 207). Hipparion ingenuum (p. 195). Megatherium mirabile (p. 37). Alligator (or Crocodylus) sp. indet.

The following vertebrates have been collected east of Williston, in the place mentioned in Dall’s report of 1892, on page 129, as Mixon’s:

Gomphotherium floridanum (p. 121). Procamelus major (p. 224). Teleoceras proterus (p. 211). Hipparion ingenuum (p. 196). Hipparion plicatile (p. 196). Thinobadistes segnis (p. 37). Manatus antiquus?. Pseudemys cælata. Atractosteus lapidosus.

The list from the locality east of Newberry (Hallowell’s place of Dall’s report) is rather short. _Equus littoralis_, _Odocoileus osceola?_, _Hipparion_ sp. indet., and _Parahippus_ sp. indet. have been reported (Rep. Florida Geol. Surv., vol. V, p. 58; vol. VIII, pp. 42, 94). At Neals, Alachua County. _Tapirus terrestris?_, _Gomphotherium floridanum_, and _Hipparion_ sp. indet. have been collected (Sellards as cited). At Juliette, same county, _Gomphotherium floridanum_ has been secured, and at Hernando the same species; also _Hipparion_ sp. indet. and _Procamelus_ sp. indet. (Sellards Florida Geol. Surv., vol. V, p. 58). Along Santa Fe River, in the Buttgenbach mines, 6 miles north of Wade, have been found teeth of _Equus_ and a tooth of _Bison_.

At Dunnellon, about 25 miles south of Williston, from the phosphate mines along the Withlacoochee River, have been obtained fossil vertebrates so similar to those found in the Alachua clays that Sellards concluded to unite his Dunnellon formation and the Alachua clays into one to be called the Alachua formation (6th Ann. Rep. Florida Geol. Surv., p. 161). The list of vertebrates found at and about Dunnellon is as follows, including the species dredged in Withlacoochee River:

Megalonyx sp. indet. (p. 38). Chlamytherium septentrionale (p. 138). Ursus sp. indet. Felis sp. indet. *Gomphotherium floridanum (p. 122). Mammut americanum (p. 122). Elephas imperator (p. 162). Trichechus manatus. *Parahippus sp. indet. (p. 196). *Hipparion plicatile (p. 196). Equus leidyi (p. 196). Tapirus sp. indet. (p. 207). *Aphelops longipes (p. 211). *Procamelus minor (p. 225). Odocoileus osceola (p. 233). Bison sp. indet. (p. 263).

The species marked by an asterisk are regarded by Doctor Sellards and others as belonging to the Miocene or Pliocene (8th Ann. Rep. Florida Geol. Surv., p. 94). See also Sellards, 1913 (5th Ann. Rep. Florida Geol. Surv., p. 58; 8th Rep., p. 104).

On the basis of the fossil vertebrates it can hardly be denied that the Alachua clays and the phosphate mines at Dunnellon are of the same geological age. According to Sellards, the formation belongs to the upper Miocene or to the lower Pliocene. Merriam (Bull. Dept. Geol. Univ. Cal., vol. X, p. 439) refers it to the Pliocene. Although there is present a strong palæontological element which represents the Pleistocene, the reference of the formation to the late Miocene or early Pliocene has seemed to be required by the presence of _Gomphotherium_, _Procamelus_, _Teleoceras_, and _Hipparion_. The Pleistocene species are usually accounted for on the supposition that they are intrusions from more recent deposits.

A figure from Sellards (Geol. Surv. Florida, vol. VII, p. 53), only slightly modified is intended to show the relation of the phosphate-bearing formations to those underlying them (fig. 21).

It is worth our while to consider whether or not the reference of the Alachua formation to the Miocene or early Pliocene is required by palæontological evidence. _Gomphotherium_ is characterized by having molar teeth which on abrasion at one or both ends of each crest, present a trefoil pattern of the enamel; also by having a band of enamel on each of the upper tusks. Now, teeth having the same structure are not uncommon in deposits of undoubted Pleistocene age in Kansas and Texas. That the animals possessing these teeth had tusks with enamel bands is not known, but it is quite possible that such enamel bands were present.

1. Georgia-Florida State line.

2. Suwannee River.

3. Lake City.

4. Santa Fe River.

5. Withlacoochee River.

6. Lakeland.

7. Arcadia.

8. Caloosahatchee River.

9. Gulf Coast.

_a_ Upper Oligocene phosphatic marls.

_b_ Ocala limestone.

_c_ Hard rock phosphate.

_d_ Bone Valley formation.

_e_ Pleistocene deposits (Pliocene and Pleistocene of Sellards).

The genus _Hipparion_ is not confined to the Tertiary. Teeth have been discovered in the Aftonian of Iowa (Hay, Geol. Surv. Iowa, vol. XXIII, p. 150) and in Missouri (op. cit., p. 149). The writer has described a species of the genus, _Hipparion cragini_, collected by Professor Cragin in the Sheridan beds in Kansas (Kansas Univ. Sci. Bull., vol. X, p. 42).

One may be justified in suspecting that _Procamelus_ lived on into the Pleistocene. Not only has it been found associated with Pleistocene fossils in five places in Florida—Archer, Williston, Dunnellon, Hernando, and Ocala—but it has been met with in possible Pleistocene deposits (the Idaho formation) in Idaho, which furnishes _Equus_, _Cervus_, _Castor_, and _Stegomastodon mirificus_ (the type of which belongs in the Sheridan beds). Furthermore, the writer has had occasion to describe a collection of fossils, believed to belong to the early Pleistocene, which was obtained at Anita, Coconino County, Arizona. Among these fossils are two species of _Procamelus_ much like those described by Leidy from the Alachua formation (Proc. U. S. Nat. Mus., vol. LIX, pp. 622–626). The writer believes that the genus _Procamelus_ persisted into the early Pleistocene.

Two species of rhinoceros have been collected in the Alachuan formation, _Teleoceras proterus_ Leidy and _Aphelops longipes_ Leidy. Both occurred at Archer, while _T. proterus_ was found near Williston and _A. longipes_ at Dunnellon. A rhinoceros has been discovered in the Idaho formation, with the Pleistocene species named above in connection with _Procamelus_ of these beds. In Oregon Cope made a collection which has been examined by Dr. W. D. Matthew (Bull. Amer. Mus. Nat. Hist., vol. XVI, p. 321). Here again _Teleoceras_ was supposed to have been found with _Hipparion_, camels belonging to _Camelops_ (or _Procamelus_), _Elephas_, and _Equus_. Matthew thought that there had happened, either before the fossils were collected or afterwards, a mingling of elements of two distinct faunas.

To the writer it seems improbable that the commingling of _Procamelus_ and the rhinoceroses with Pleistocene forms should occur thus accidentally so often and at such widely removed localities. It appears more probable that these Tertiary genera did not become extinct so early as has been supposed and that the association was not a secondary one. The association is what might be expected in collections made in deposits of the earliest Pleistocene.

It must not be forgotten in these discussions that the Pleistocene genera and species with which the collections in question are being compared are those of the so-called _Equus_ beds, which appear to represent the fauna of the first interglacial stage. This, however, was preceded by the Nebraskan, the first glacial, which probably occupied a long period of time; possibly it was half as long as all the rest of the Pleistocene (Chamberlin and Salisbury, Geology, vol. III, p. 383). About the vertebrate life of this long stage we know as yet very little. The writer is quite convinced that the Idaho formation and the Alachua, or Bone Valley, belong to the earliest Pleistocene.

_Marion County._—In a fissure in the limestone-rock quarry at Ocala there has been found an important collection of vertebrates. The following list is thought to include all that have been reported:

Trucifelis floridana. Sylvilagus sp. indet. Elephas columbi (p. 158). Bison sp. indet. Odocoileus sp. indet. (p. 233). Procamelus minimus (p. 224). Tapirus sp. indet. (p. 207). Equus leidyi (p. 196). Dasypus sp. indet. (p. 38). Terrapene formosa. Testudo distans. T. incisa. T. ocalana.

A part of this list was published by Sellards in 1916 (8th Ann. Rep. Florida Geol. Surv., p. 103). The tortoises were described in the same volume.

Inasmuch as _Trucifelis floridana_ has been found in the Pleistocene at Vero, Florida, one may safely regard the specimen found at Ocala as also of Pleistocene age. All of the other mammals are admitted to be of Pleistocene age except _Procamelus minimus_. The fissure may have been open during some part of the Nebraskan stage.

_Volusia County._—At Daytona, situated on the east coast, therefore on the youngest terrace, remains of _Mammut americanum_ (p. 122) have been found. At DeLand there has been recovered the skull of a dolphin which Sellards (8th Ann. Rep. Florida Geol. Surv., p. 107, plate XIV) has described as _Globicephalus bœreckii_ (p. 20). It was found at a depth of 10 feet, in sands which overlie Pliocene shell marls. The sands are regarded as belonging probably to the Pleistocene. DeLand is on the Tsala Apopka terrace. At a depth of 10 feet there was reached the supposed marine base of this terrace.

_Orange County._—As stated on page 196, a tooth of an extinct horse was found somewhere in the county.

_Pinellas County._—On the western shore of Tampa Bay (p. 159), near St. Petersburg, at Indian Rock, a tooth of _Elephas columbi_ was found.

_Hillsboro and Manatee Counties._—The region around Tampa Bay is important because of the wealth of vertebrate fossils dredged up by the collectors of phosphate rock from the beds of Hillsboro, Alafia, and Manatee Rivers. Unfortunately, few accurate records have been kept of localities and conditions of occurrence of the fossils, and we usually know only that a collection was made in a certain river, perhaps not so much as that. For that reason it is concluded to group together all the fossils regarded as Pleistocene and known to have been found in Hillsboro, Manatee, and Sarasota Counties. In order to indicate as far as possible the localities, the names of the species are followed by contractions which apply as follows.

A, Alafia River. E, Ellenton, on Manatee River. Hc, Hillsboro County. Hr, Hillsboro River. Ls, Little Sarasota Bay. Ps, Palma Sola. M, town of Manatee. P, Palmetto. S, 8 miles southeast of Sarasota. T, around Tampa Bay. Wb, White Beach, on Sarasota Bay.

_List of Pleistocene vertebrates found in Hillsboro, Manatee, and Sarasota Counties._

Homo sapiens, Ps. Elephas imperator, P (p. 164). E. columbi, T, S, Ps (p. 159). E. primigenius, Ps (p. 145). Mammut americanum, T, He, A (p. 123). Bison latifrons, Ps (p. 263). B. sp. indet., T, A, P, Ps (p. 263). Odocoileus sp. indet., P, Ps (p. 233). Tapirus sp. indet., T, A (p. 208). Tagassu lenis?, Ps (p. 222). Equus complicatus, A, Wb, Ps (pp. 196, 197). E. leidyi, A. P, Wb, Ps, S (pp. 196, 197). E. littoralis, M, Ps (p. 197). Chlamytherium septentrionale, Hr, Wb, S (p. 38). Testudo crassiscutata?, A. T. ocalana, He. Terrapene putnami, A. Deirochelys floridana, He. Trachemys euglypha, E. T. jarmani, Hc. T. sculpta, Hc. Pseudemys extincta, Hc. P. cælata, Ps. Platypeltis ferox, Hc.

The bones of man belonged to the skull and are as completely fossilized as the bones of a horse and are wholly free from organic matter.

Among the mammals of this list there are no genera and few species that have not been found in the Pleistocene at many places in the United States. The presence of _Elephas imperator_ and three species of _Equus_ and _Chlamytherium_ apparently indicate Pleistocene of about Aftonian times.

From Palma Sola, Manatee County, there have been sent to the U. S. National Museum by Mr. Charles T. Earle many specimens of fossil vertebrates, found at various times washed up on the beach. Some belonged evidently to deposits older than the Pleistocene, probably to Miocene, and included teeth of sharks, a beak of a platanistid porpoise, and a lower tooth of a sirenian, _Metaxytherium floridanum_. Other specimens, as bones of a camel, parts of the shells of tortoises, alligator or crocodile teeth and bones are of uncertain age. Ten species of the list are referred to the Pleistocene. All of the teeth are isolated, but many are well preserved and little water-worn. The bones are mostly fragmentary, some worn, some not.

_Polk County._—On page 159 is an account of a tooth of an elephant. _Elephas columbi_, reported as being found at Kingsford, Polk County, under 19 feet of phosphate rock and sand. It may belong to _E. imperator_. On page 196 is detailed the finding of several teeth of _Equus_ in the phosphate mines of Kingsford. The species _E. leidyi_ and _E. littoralis_ are recognized. Unless these elephant and horse-teeth had been incorrectly reported or had been secondarily introduced into the phosphate beds, they are, in the writer’s opinion, to be referred to the first glacial stage, the Nebraskan. Dr. W. H. Dall has somewhere reported the finding of tusks at Bartow; these were supposed to have belonged to _Elephas columbi_ (p. 180). At Nichols the large land-tortoise _Testudo hayi_ Sellards has been recovered from a phosphate mine. From phosphate mines at Brewster has been secured the following list of vertebrates, obtained from Dr. Sellards’s reports (Florida Geol. Surv., vol. VII, pp. 100, 106, 108; vol. VIII, pp. 95, 96, 98, 100).

Gomphotherium floridanum? (p. 123). Mammut progenium (p. 123). Hipparion minor (p. 197). Procamelus minor? Teleoceras or Aphelops sp. indet. (p. 211). Agriotherium schneideri. Tomistoma americanum.

All of this list are referred by Sellards to the upper Miocene or lower Pliocene. The writer regards them as belonging to the first stage of the Pleistocene.

From a phosphate pit at Christina, Sellards (Florida Geol. Surv., vol. VII, p. 106, fig. 35) has reported a tooth of an undetermined species of Gomphotherium.

Sellards (Florida Geol. Surv., vol. VII, pp. 72, 110) has reported the collection of remains of _Hipparion_ sp. indet. and of _Teleoceras proterus_ (p. 211) from phosphate mines at Mulberry. In the U. S. National Museum are undetermined remains of _Gomphotherium_ from the same place, sent in by Matson.

_Brevard County._—In the Hopkins drainage canal at Eau Gallie have been found remains of _Equus complicates_ (p. 196) and _Elephas columbi_ (p. 159).

_Zolfo, Hardee County._—At Zolfo, near the border of the Bone Valley area, have been found _Megatherium_ (p. 38) and _Elephas columbi_ (p. 160).

_De Soto County._—With one exception, apparently, fossil vertebrates have been discovered in De Soto County only in deposits along Peace Creek. The exception is a place called Tourner’s or Turner’s, on Caloosahatchee River. The elephant found there will be considered among the fossils found in Lee County. At Calvenia, at the entrance of Charlie Apopkee Creek into Peace Creek, _Equus leidyi_ (p. 198) has been secured.

Most of the fossils found below Calvenia are accredited to Arcadia. According to Leidy (Trans. Wagner Inst., vol. II, p. 19), those of his list were found on a sand-bar at Arcadia; but certainly others have been taken from phosphate rock dredged both above and below the town. As complete and as accurate a list as the writer has been able to prepare is here presented.

Peace Creek, or Peace River, has been the source of many fossil vertebrates, the greater part of them obtained at or near Arcadia. Most of the species were described by Joseph Leidy in 1889 (Trans. Wagner Free Inst., vol. II, pp. 19–31). The region was examined by Dr. W. H. Dall, whose report was published in 1892 (Bull. 84, U. S. Geol. Surv., pp. 128–133). He referred the bed bearing vertebrate fossils to the Pliocene. Cope (in Dall’s report, p. 130) regarded them as equivalent to the _Equus_ beds of the Great Plains, or between these and the Loup Fork. Sellards (Florida Geol. Surv., vol. VII, pp. 78–83) places the formation in the Pleistocene.

_List of fossil vertebrates found in Peace Creek at or near Arcadia._

Elephas imperator (p. 164). E. columbi (p. 160). Mammut americanum (p. 124). Bison sp. indet. (p. 264). Odocoileus osceola? (p. 233). Tapirus terrestris (p. 208). Equus leidyi (p. 199). E. littoralis (p. 199). Hipparion ingenuum (p. 199). Delphinid sp. indet. Trichechus antiquus. Glyptodon petaliferus (p. 39). Glyptodon rivipacis (p. 40). Chlamytherium septentrionale (p. 40). Megalonyx jeffersonii (p. 39). Alligator mississippiensis?. Testudo crassiscutata. T. obtusa. Trachemys euglypha. Macrochelys floridana. Chætodipterus faber. Diodon sp. indet. Myliobatis sp. indet. Galeocerdo sp. indet. Isurus sp. indet.

Of all the genera and species of mammals and reptiles appearing in the list, there is none that it is necessary to suppose was derived from Pliocene deposits, or even from those of a Pleistocene stage earlier than the first interglacial. The marine fishes and sharks have been derived possibly from the Arcadia marls. On the other hand, the presence of _Elephas imperator_, the species of _Equus_, _Hipparion_, _Glyptodon_, _Chlamytherium_, and the gigantic tortoise _Testudo crassiscutata_ furnishes evidence that the age was about that of the _Equus_, or Aftonian, beds of the Great Plains.

_St. Lucie County._—At Fellsmere, a place near the northern border of the county and about 10 miles west of Indian River, teeth of both _Elephas columbi_ (p. 159) and _Mammut americanum_ (p. 122) have been found, in the construction of drainage canals.

The most important locality for Pleistocene fossils in St. Lucie County, one may say in the whole State, is Vero. The topographical, geological, and palæontological conditions found here are described in the Eighth and Ninth Annual Reports of the Florida Geological Survey. Papers on the subject may be found also in the Journal of Geology for January 1917 and for October 1917; also in the American Anthropologist for the first and second quarters of 1918. Besides the large number of species of vertebrates found here, the interest is heightened by the fact that, associated with these, are human bones and objects of human manufacture. Through the valley of an insignificant stream was dug a large drainage canal, the construction of which brought to light vertebrate bones and teeth. Three beds of Pleistocene materials were exposed. At the bottom is found a bed of marl filled with marine mollusks and which is the geological equivalent of the coquina rock at St. Augustine. The same deposit is found in various places along the coast and has received from Dr. Sellards the name Anastasia formation. Above this lies a stratum composed mostly of sand, but containing also some muck. In the discussion of the locality this bed is designated as No. 2, the marl being No. 1. No. 2 has a thickness of about 2 feet. It in turn is overlain by No. 3, which consists mostly of vegetable matter and sand. It is called also the muck-bed. In places the muck is replaced by a bed of marl, which here and there may become pretty firmly consolidated. The thickness of No. 3 is about 2 or 3 feet. Vertebrate fossils are found in both No. 2 and No. 3. It is the purpose of the author first to present lists of the fossils which have been found in each of the upper beds, beginning with the stream of sand, No. 2.

_List of fossil vertebrates found at Vero in stratum No. 2._

Trucifelis floridanus. Felis veronis. Ænocyon ayersi. Vulpes palmaria. Lutra canadensis. Procyon lotor. Cryptotis floridana. Blarina brevicauda peninsulæ. Sylvilagus sp. indet. Neofiber alleni. Hydrochœrus robustus. Sigmodon sp. indet. Elephas columbi (p. 159). Mammut americanum (p. 122). Bison sp. indet. (p. 263). Odocoileus osceola? (p. 233). Camelops? sp. indet. (p. 235). Tagassu lenis (p. 222). Tapirus haysii? (p. 208). T. veroensis (p. 208). Equus complicatus (p. 199). E. leidyi (p. 199). E. littoralis (p. 199). Dasypus sp. indet. (p. 38). Chlamytherium septentrionale (p. 38). Megalonyx jeffersonii (p. 38). Mylodon harlani (p. 39). Didelphis virginiana. Jabiru weillsi. Cathartes aura. Querquedula floridana. Herodias egretta. Alligator mississippiensis. Testudo sellardsi. Terrapene innoxia. Chelydra laticarinata.

_List of fossil vertebrates found at Vero, in stratum No. 3._

Lynx ruffus floridanus. Canis riviveronis. Canis sp. indet. Vulpes palmaria. Lutra canadensis. Ursus floridanus. Procyon lotor. Scalopus aquaticus australis. Sylvilagus palustris. Neofiber alleni. Neotoma floridana. Sigmodon hispidus. Oryzomys palustris. Elephas columbi (p. 159). Mammut americanum (p. 122). Odocoileus osceola? (p. 233). O. sellardsiæ (p. 233). Bison sp. indet. (p. 263). Tagassu lenis (p. 222). Tapirus haysii? (p. 208). Equus littoralis? (p. 199). E. leidyi? (p. 199). Chlamytherium septentrionale (p. 38). Dasypus sp. indet. (p. 38). Didelphis virginiana. Ardea sellardsi. A. herodias. Ardea? sp. indet. Herodias egretta? Aluco pratincola. Cathartes aura septentrionalis. Larus vero. Larus? sp. indet. Mycteria americana? Alligator mississippiensis. Crotalus adamanteus. Farancia abacura. Drymarchon corais couperi. Gopherus polyphemus. Terrapene antipex. T. innoxia. Pseudemys floridanus persimilis. Trachemys? nuchocarinatus. Chelydra sculpta. Chelonia mydas. Caretta caretta. Siren lacertina. Amphiuma means. Caranx hippos. Caranx sp. indet. Amiatus calvus. Lepisosteus platystomus. Aëtobatis narinari.

Besides those remains which are to be assigned with certainty to one or the other or both of the strata, there are a few others about whose place in the deposit there is uncertainty:

Testudo luciæ. Gopherus præcedens. Trachemys bisornata. T. sculpta.

At a point about 3 miles west of Vero, a lower jaw of _Elephas imperator_ (p. 163) was found in the bank of the drainage canal. It was embedded in a matrix of brown sand which rests upon the stratum of marine shell marl.

The list of mammals found in stratum No. 2 shows that there are 29 species and that 21 of these are extinct. This high proportion of species no longer existing is of itself enough to show that the deposit is an old one. Again, such species as _Elephas imperator_ and camels occur in the glaciated region only in Aftonian beds, and outside of the glaciated region only in those which are quite certainly of approximately the same age.

In the list of species found in stratum No. 3 there are 25 mammals, of which 12 species are extinct. These form, therefore, 48 per cent of the whole, indicating apparently a more recent geological time, perhaps about the Sangamon stage. It is true that the geologists hold that there has been continuous deposition and that no interval elapsed between the laying down of No. 2 and No. 3. In a region so near to the level of the sea, where the streams are small and short and have little fall, deposition must have gone on with extreme slowness; hence there may have been no period when deposition ceased. Apparently, too, there was a time when the region was somewhat lower than at present and salt water came up the stream as far as the locality where the fossils are found. The presence of _Chelonia mydas_, _Caretta caretta_, the two species of _Caranx_ and _Aëtobatis narinari_ may thus be explained.

The fresh-water and terrestrial mollusks of stratum No. 2 were submitted to Dr. Paul Bartsch, of the U. S. National Museum, who has reported on them (Rep. Florida Geol. Surv., vol. VIII, p. 144). He lists 29 species, all living.

The marine mollusks found in the stratum called No. 1, and which the writer refers to the first glacial stage, have been studied by Mr. W. C. Mansfield, of the U. S. Geological Survey (Rep. Florida Geol. Surv., vol. IX, pp. 78–80). Seventy-four species are specifically determined, and of these 61 are identical with living forms. Three or four species are possibly extinct. There is no question that the deposit belongs to the Pleistocene.

Nearly all of the plants were found in the bed designated as No. 3, the upper or muck-bed. These were studied by Dr. Edward W. Berry, of the Maryland Geological Survey. His report, published in 1917 (Rep. Florida Geol. Surv., vol. IX, pp. 19–33), states his conclusion that the plants belong to the late Pleistocene, either the Peorian or the Late Wisconsin. It may be stated that Dr. Berry adopts the theory that the terraces supposed to be found along the Atlantic Coast were formed during stages of submergence beneath the sea, the lowest one late in Pleistocene time.

_Lee County._—The whole of Lee County is occupied by Pleistocene deposits which form a part of the Pensacola terrace. Naturally the Pleistocene is overlain, generally, at least, by accumulations of Recent materials, and it may not always be easy to distinguish the one from the other. So far as the writer knows, all the vertebrate fossils discovered in this county have been collected along Caloosahatchee River above Fort Myers. The geology of this river has been described by Heilprin (Trans. Wagner Inst., vol. I), Dall (Bull. 84, U. S. Geol. Surv., pp. 142–145), Matson and Sanford (Water Supply Paper 319, pp. 134–138), Sellards (2d Ann. Rep. Florida Geol. Surv., p. 123, 6th Ann. Rep. Florida Geol. Surv., pp. 41–46). The Pleistocene is underlain by Pliocene marls and hard and soft limestones and consists of beds of muck, marl, and sand of little thickness. At Labelle it is said (Sellards, 2d Ann. Rep., p. 126) that there is a fossiliferous Pleistocene marl covered by 3 feet of sandy loam. The following seem to be the species which have been found in the Pleistocene in this region:

Elephas imperator (p. 163). Equus leidyi (p. 199). E. sp. indet. (p. 199). Bison latifrons (p. 264). Mylodon harlani (p. 40). Testudo obtusa?. Trachemys bisornata. T. sculpta.

The presence of _Elephas imperator_ is an indication that the deposits belong to the early part of the Pleistocene. None of the species appear to indicate an older stage than the Aftonian.

_Dade County._—Sellards (8th Ann. Rep., p. 106) records that some fragmentary remains of a proboscidean had been found in Miami River, Dade County.

_Palm Beach County._—On page 105 of the report just cited, Sellards stated that _Elephas columbi_ (p. 160), _Mammut americanum_ (p. 123), _Equus complicatus_ (p. 200), and _Bison_ sp. indet. (p. 264) had been found in the Palm Beach Canal, constructed to drain the Everglades.

At some unknown point in the Everglades, possibly in Lee County, there was found many years ago a tooth of an elephant which the writer believes belonged to _Elephas imperator_, already mentioned on page 163. It was formerly reported as _E. columbi_.

ALABAMA.

An account of the Quaternary formations of Alabama may be found in Eugene A. Smith’s “Report on the Geology of the Coastal Plain of Alabama.” This was published in 1894, and the part pertaining to the Pleistocene is found on pages 28 to 65. Along the coast Smith recognized the presence of a formation which he called the Biloxi. The upper part of this was regarded as belonging to the Recent, while the lower portion was thought to be the equivalent of Hilgard’s Port Hudson, those deposits numbered 1 to 4 in the section shown on page 387, under Geology of Mississippi. The thickness of the Port Hudson is given as about 100 feet. Borings revealed the presence of shells and lignitized wood.

Along the rivers which traverse the Coastal Plain are found three terraces. The first or lowest is that which is subject to annual overflow. The second terrace, “the second bottom,” occurs along most of even the smaller streams of the Coastal Plain. It may be as much as a mile wide. The height above low water may vary from 10 to 15 feet in the lower courses of the rivers to 60 feet farther up stream. Near water-level a blue clay is frequently found which contains stumps, roots, and other remains of vegetation, often well preserved. Smith concluded that this second terrace was the substantial equivalence in time to the Port Hudson.

Smith presents a geological section taken along Black Warrior River, in Hale County, 150 miles above Mobile. The section included about 50 feet. As caving went on, stumps and logs were frequently brought into view. Similar sections were found on Coosa River, above Montgomery, and on Alabama River, 50 miles above Mobile.

The third terrace is found at elevations of from 50 to 100 feet above the second. It is sometimes 3 miles or more in width.

In his paper on the Citronelle formation (Prof. Paper U. S. Geol. Surv., 98, pp. 167–208), Matson discusses briefly (pp. 189–190) the Pleistocene of the area studied by him. This extends from the western end of Florida to Mississippi River. Here he recognized four terraces, from the youngest to the oldest, the Pensacola, the Hammond, the Port Hickey, and the St. Elmo. The St. Elmo merges into the Natchez formation, which Matson, quoting Chamberlin and Salisbury, regarded as sub-Aftonian. The Port Hickey terrace is stated to take its name from a locality on the Mississippi River where the typical materials of the Port Hudson formation are exposed. The Port Hickey terrace may, as suggested by Matson, be of post-Iowan age. Naturally, these correlations require confirmation.

Berry has described fossil plants (Amer. Naturalist, vol. XLI, pp. 689–697; Amer. Jour. Sci., ser. 4, vol. XXIX, pp. 387–398) which were found along Chattahoochee River, not far below Columbus, Georgia; on Warrior River, up to 356 miles above Mobile. Pleistocene deposits must occur along all the larger streams still farther north, and these deposits will yield in time bones and teeth of vertebrated animals.

Notwithstanding the considerable area of Pleistocene deposits discovered in Alabama, the number of species of vertebrates met with is remarkably small. On page 40 is recorded the finding of _Megalonyx jeffersonii_ somewhere about Tuscumbia. At Newbern, Hale County, have been found an incisor tooth of a horse (p. 200) and a molar of a bison (p. 264). At Bogue Chitto, Dallas County, have been collected _Equus leidyi_ (p. 200), _Mammut americanum_, and _Elephas imperator_. The last species indicates that the deposits probably belong in the Aftonian. The writer knows of no other localities in the State where vertebrate fossils of the Pleistocene have been obtained.

MISSISSIPPI.

(Text-figure 22.)

The geological history of the lower part of the Mississippi Valley during Quaternary times appears to be particularly difficult to understand and at present is far from being unraveled. It is easy to see that such a region will offer great difficulties. Here debouches into the ocean a majestic river which drains not only the glaciated portions of the United States from western New York to northwestern Montana, but the larger part of the region south of this from the Blue Ridge to the Rocky Mountains, and brings down every year enormous quantities of sand and silt, which are dropped partly on its flood-plain, but mostly near its mouth. Through the ages during which this has been proceeding, this river has been ever changing its bed, sometimes eroding away one bank, sometimes the opposite one; so that its flood-plain is, in most places below the mouth of the Ohio, many miles wide, varying, according to Russell (“Rivers of North America,” 1898, p. 267) from 5 miles to 80 miles in width. During the Quaternary there have been also elevations and subsidences of the bed at least from Cairo northward, as a result of which at one epoch the current was hastened and the valley cut out deeper; at another the current was checked, the channel clogged up, and the river forced to seek a new channel or even new temporary or permanent outlets to the Gulf (E. A. Smith, Geol. Surv. Alabama, 1894, pp. 30–34).

To get a correct idea of the Pleistocene geology of the lower Mississippi region, one must understand the situation at the beginning of this epoch. I. C. Russell, on page 267 of his work just quoted, calls attention to the differences displayed by the valley of the river within the glaciated region and that south of it. South of the mouth of Ohio River the wide flood-plain of the Mississippi lies from 300 to 500 feet below the general level of the bordering uplands. He states further that the hard rock bottom of the valley is only imperfectly known, but that the records of wells and borings show that an ancient valley has been filled with alluvium to a depth of at least 100 or 200 feet in its northern part and to an increasing depth southward. If to this thickness, given by Russell, we add the depth, 300 to 500 feet, which the flood-plain occupies below the bordering uplands, we get a measure of the depth of the great trench which once existed where now lies the flood-plain of Mississippi River. In his paper on the underground waters of southern Louisiana (Bull. 1, Louisiana Geol. Surv., 1905, p. 42, plate II) Harris presents the record of the Fabacher well, which was bored at New Orleans. At a depth of about 1,200 feet fossil remains were brought up which appeared to be of Pleistocene age. It is evident from these facts, as in the case of those obtained from the rivers of Texas, that at about the beginning of the Pleistocene, or more probably during the time of the so-called Lafayette, at the close of the Pliocene, the country east of the Rocky Mountains, at least, stood for a long time at a much higher level than at present and that, as a result of this elevation, there was an enormous general erosion of the face of the country and a great widening and deepening of the river valleys. This time of elevation was quite certainly followed by a prolonged period of depression, during which these canyon-like trenches and their tributaries, up to their last ramifications, were nearly completely refilled. This refilling must have occurred during the early stages of the Pleistocene, for in the materials are buried the bones of early Pleistocene animals. As quoted below, in considering the geology at Natchez, Chamberlin and Salisbury state that since the Natchez formation, 200 feet thick, was laid down, the trench of the Mississippi, 60 miles wide, has been excavated. One might change this expression and say that it had been re-excavated, but not to its original depth.

When we reflect that the greater part of the sediments which, during the Pleistocene epoch, were deposited at the mouth of Mississippi River and on its flood-plain from Kentucky southward, were certainly derived from the glaciated portions of its great valley, and that those regions were alternately affected by the events of five glacial and four interglacial epochs, we must conclude that corresponding deposits or phenomena of some kind exist throughout the valley. The matter is, however, so complicated that many years must elapse before a satisfactory solution will have been reached.

In his Report on the Geology and Agriculture of the State of Mississippi, 1860 (1863–65), the geologist E. W. Hilgard, on pages 5 to 46, described under the name of Orange sand a deposit which characterizes the greater part of the surface of that State. He referred this to the Quaternary and regarded it as being the southern equivalent of the northern drift. This formation is now believed to belong mostly at least to the Pliocene. Besides the Orange sand, Hilgard (op. cit., pp. 194–201) referred other formations to the Quaternary. These in order would be as follows, the latest above:

5. Modern alluvium.

4. Second bottom, or Hommock deposits.

3. Yellow loam deposits.

2. The Bluff formation.

1. Orange sand.

The Bluff formations were described as occupying a narrow belt along the borders of the Mississippi bottom in northern Mississippi and along the river itself in the southern part of the State. He stated that the fossils belonged to terrestrial species, and quoted Leidy’s list of vertebrates, already mentioned, remarking that the blue clay which furnished them was said to belong to the Bluff formation. He reported that the snails found in the Bluff formation seemed all to belong to living species. The yellow loams occupied a large part of the surface of the State, overlying the Orange sand and forming a great part of the soils of the State. The succeeding formations were found along many of the rivers.

In 1869 (Amer. Jour. Sci., vol. XLVII, pp. 331–346), Hilgard reported the results of a geological reconnaissance of Louisiana. In this he proposed the name Port Hudson group for extensive deposits of clays which were especially well displayed at Port Hudson. This formation was further described by Hilgard in 1872 (Smithson. Contrib. Knowl., vol. XXIII, No. 248, p. 5). Two geological sections taken near Port Hudson were presented, one of which is here reproduced.

_Section midway between Port Hudson and Fontana._

6. Yellow loam, sandy below 8–10 5. White and yellow hardpan 18 Orange and yellow sand, sometimes ferruginous sandstone, irregularly stratified 8–15 4. Heavy greenish or bluish clay 7 3. White indurate silt, or hardpan 18 2. Heavy green clay with porous calcareous concretions above, ferruginous below; some sticks and impressions of leaves 30 1. Brown muck with cypress stumps 3–4 White or blue clay with cypress stumps

The cypress stumps of No. 1 were numerous and well preserved.

The writer reproduces Hilgard’s geological map of the lower Mississippi region, in which is represented the distribution of the Port Hudson according to that writer’s views (fig. 22). It will be seen that it was supposed to pass eastward into the coast region of Mississippi, Alabama, and Florida. Westward from Atchafalaya River it was believed to occupy a large part of southern Louisiana and to pass into Texas and around the Gulf coast to near the Rio Grande. It will be observed that in the latter State it corresponds in a general way to what has been called by Deussen the Lissie formation.

Although Hilgard represents on his map an alluvial deposit as covering the region of the delta, a belt along the western side of the great river as far up as Cairo, and the wide tract between Mississippi and Yazoo Rivers, a study of his paper shows that he believed that much of these regions was underlain by his Port Hudson. He recognized it at Greenwood on the Yazoo, 60 miles east of the Mississippi; at Vicksburg, and at various places in the delta. Usually its upper surface occurs at about low-water level along rivers, and elsewhere is met with in digging wells. At Vicksburg it was encountered by Grant’s Army in digging his famous canal. It was believed by Hilgard that the same deposit was present at Petite Anse, overlying the Orange sand and overlain by more recent deposits.

Inasmuch as Hilgard believed that the Orange sand was laid down at the time when the northern drift was being deposited, he had to refer his Port Hudson to a later time, and this time he seemed to regard as being the epoch called by Dana the Champlain.

McGee referred the deposits of the lower Mississippi Valley (sometimes called the Mississippi embayment) to his Columbia formation (12th Ann. Rep. U. S. Geol. Surv., plate I, p. 392). This formation, in his view, had been laid down during a great subsidence of the borders of the continent and when the waters of the Gulf reached as far north as the mouth of Ohio River or beyond. He relegated Hilgard’s Orange sand to the Pliocene and recognized four phases as belonging to the Pleistocene. These were, beginning below: (1) Port Hudson; (2) Orange sand (of Safford, not that of Hilgard); (3) loess; (4) brown (or yellow) loam. Of these divisions there were really only three, for he regarded the loess as only a phase of the loam and as lying sometimes above, sometimes below the latter. He recognized the Port Hudson clays as flooring the entire flood-plain of the Mississippi from the mouth of the Ohio well toward the gulf shore. The formation was believed to be usually a low-lying one; but at Natchez (as seen by his section given on page 391) it is elevated high above the present flood-plain.

Gilbert D. Harris, geologist in charge of the geological survey of Louisiana, and Arthur C. Veatch, assistant geologist, have contributed much to our knowledge of the Pleistocene geology of the State. Reference to their works will be found in the descriptions of several fossil-bearing localities, especially in the description of Petite Anse.

Harris, in 1905 (Bull. 1, Geol. Surv. Louisiana, p. 13), expressed the conclusion that the longer the geology of southern Louisiana is studied the more futile appears the attempt to make satisfactory subdivisions in the Quaternary deposits—subdivisions that have any definite time or structural limits. He regarded it as a mistake to assign to the Port Hudson a special place in geologic time.

Chamberlin and Salisbury in 1906, as quoted below, made no mention of the Port Hudson formation; but that part of it supposed to be found at Natchez was evidently included in their Natchez formation.

Inasmuch as Petite Anse and Natchez have furnished more species of fossil vertebrates than any other localities in their States, and likewise human relics supposed to be of equal age with the extinct mammals, these places will receive especial attention.

Natchez is the most important locality in Mississippi as regards Pleistocene vertebrate palæontology. So far as the writer knows the first mention of the occurrence of vertebrate fossils here was a note by Dr. G. Troost in 1835 (Trans. Geol. Soc. Penn., vol. I, p. 143), who stated that he had in his possession a tooth of a mastodon found at Natchez.

In 1845 (Proc. 6th Meet. Assoc. Amer. Geologists and Naturalists, pp. 77–79), Dr. M. W. Dickeson, of Natchez, read a paper entitled “On the Geology of the Natchez Bluffs,” in which he distinguished 22 several beds. These were said to be of varying thickness and distinctly marked, but all composed of various colored clays and sands, and containing numerous organic remains, embedded wood, and detrital matter. Probably by far the greater part of these beds were of subordinate importance and do not appear to have been noted since that time. Beneath the surface soil Dickeson recognized a mass of yellow loam 20 to 30 feet in thickness, exceedingly fine and free from gravel. In this had been found shells of _Helix_ and scattered bones of mastodons. Below this came a bed of ferruginous sands and gravels 4 feet thick. This was succeeded below by what he called the mastodon bed, in which Dickeson had detected remains of more than 30 individual mastodons. The thickness of this was not given. The next stratum, his No. 6, was a fine clay of blue color, from 12 to 15 feet thick. In this and his No. 22, an ash-colored clay, at low-water mark, he discovered remains of what has since proved to be _Megalonyx jeffersonii_. The localities where his fossils were found were not given with exactness.

At a meeting of the Academy of Natural Sciences, October 6, 1846 (Proc., etc., vol. III, p. 107), Dickeson exhibited a large collection of fossil bones obtained by him in the vicinity of Natchez. Among these were the head and lower jaw of the _Megalonyx_ already mentioned. He stated that the stratum that contains these organic remains is a tenacious blue clay that underlies the diluvial drift east of Natchez and which diluvial deposit abounds in bones and teeth of the _Mastodon giganteus_. Associated with the megalonyx were remains of bear, bison, deer, and horse. The collection was more notable because of the presence of a part of a human innominate bone. Dickeson affirmed that this had been taken out of the blue clay about 2 feet below three associated skeletons of the megalonyx; and it is further stated to have accorded in respect of color, density, etc., with those of the megalonyx and other associated bones. This bone is now in the collection of the Academy of Natural Sciences, Philadelphia.

In 1846 the English geologist, Sir Charles Lyell, studied the geology of the region at Natchez (Second Visit to U. S. N. A., ed. 2, vol. II, pp. 194–201). With him were Dr. Dickeson and B. L. C. Wailles, afterwards State geologist of Mississippi (Wailles, Agric., Geol. Miss., 1854, p. 213). In the yellow loam of the bluffs Lyell recognized loess deposits, from their resemblance to those of the Rhine. These he estimated to occupy the upper 60 feet of the bluff, and in them were found 20 species of land-snails, all yet living. He reported that this loess sometimes passed into a lacustrine deposit which contained shells of _Lymnæa_, _Planorbis_, _Paludina_, _Physa_, and _Cyclas_, and that with the land-snails had been found, at different depths, remains of the mastodon, while in clay under the loam (meaning evidently the loess) and above the sand and gravel, entire skeletons of the megalonyx had been met with, associated with bones of the horse, bear, stag, ox (_Bison_). Lyell noted especially the recent development of deep ravines. One of these, called the Mammoth Ravine, had been formed, he was assured, within the preceding 35 years. Its length was 7 miles and its depth 60 feet. In this ravine was found the human innominate bone referred to above. He was shown this bone, and states that Dr. Dickeson was persuaded that the bone had been taken out of the clay underlying the loam (loess). This indicates that Dickeson himself did not take out the human bone. Lyell thought that, like most of the other fossils, it had been picked up in the bed of the stream, which would simply imply that it had been washed out of the cliffs, and that it may have been dislodged from some Indian grave near the top. He (p. 197) stated that the place where the bone was found was 6 miles from Natchez. The reader may consult further Lyell’s account of his observations at Natchez in