CHAPTER III.

_ON THE NATURE OF METAMORPHOSES._

In the preceding chapters we have considered the life history of insects after they have quitted the egg; but it is obvious that to treat the subject in a satisfactory manner we must take the development as a whole, from the commencement of the changes in the egg, up to the maturity of the animal, and not suffer ourselves to be confused by the fact that insects leave the egg in very different stages of embryonal development. For though all young insects when they quit the egg are termed "larvæ," whatever their form may be (the case of the so-called Pupipara not constituting a true exception), still it must be remembered that some of these larvæ are much more advanced than others. It is evident that the larva of a fly, as regards its stage of development, corresponds in reality neither with that of a moth nor with that of a grasshopper. The maggots of flies, in which the appendages of the head are rudimentary, belong to a lower grade than the grubs of bees, &c., which have antennæ, mandibles, maxillæ, labrum, labium, and, in fact, all the mouth parts of a perfect insect.

The caterpillars of Lepidoptera are generally classed with the vermiform larva of Diptera and Hymenoptera, and contrasted with those of Orthoptera, Hemiptera, &c.; but, in truth, the possession of thoracic legs places them, together with the similar larvæ of the Tenthredinidæ, on a decidedly higher level. Thus, then, the period of growth (that in which the animal eats and increases in size) occupies sometimes one stage in the development of an insect, sometimes another; sometimes, as for instance in the case of _Chloëon_, it continues through more than one; or, in other words, growth is accompanied by development. But, in fact, the question is even more complicated than this. It is not only that the larvæ of insects at their birth offer the most various grades of development, from the grub of a fly to the young of a grasshopper or a cricket; but that, if we were to classify larvæ according to their development, we should have to deal, not with a simple case of gradations only, but with a series of gradations, which would be different according to the organ which we took as our test.

Apart, however, from the adaptive changes to which special reference was made in the previous chapter, the differences which larvæ present are those of gradation, not of direction. The development of a grasshopper does not pursue a different course from that of a butterfly, but the embryo attains a higher state before quitting the egg in the former than in the latter: while in most Hymenoptera, as for instance in Bees, Wasps, Ants, &c., the young are hatched without thoracic appendages; in the Orthoptera, on the contrary, the legs are fully developed before the young animal quits the egg.

Prof. Owen,[16] indeed, goes so far as to say that the Orthoptera and other Homomorphous insects are, "at one stage of their development, apodal and acephalous larvæ, like the maggot of the fly; but instead of quitting the egg in this stage, they are quickly transformed into another, in which the head and rudimental thoracic feet are developed to the degree which characterizes the hexapod larvæ of the _Carabi_ and _Petalocera_."

I quite believe that this may have been true of such larvæ at an early geological period, but the fact now appears to be, so far at least as can be judged from the observations yet recorded, that the legs of those larvæ which leave the egg with these appendages generally make their appearance before the body-walls have closed, or the internal organs have approached to completion. Indeed, when the legs first appear, they are merely short projections, which it is not always easy to distinguish from the segments themselves. It must, however, be admitted, that the observations are neither so numerous, nor in most cases so full, as could be wished.

Fig. 30 represents an egg of a May-fly (_Phryganea_), as represented by Zaddach in his excellent memoir,[17] just before the appearance of the appendages. It will be seen that a great part of the yolk is still undifferentiated, that the side walls are incomplete, the back quite open, and the segments merely indicated by undulations. This stage is rapidly passed through, and Zaddach only once met with an egg in this condition; in every other specimen which had indications of segments, the rudiments of the legs had also made their appearance, as in Fig. 31, which, however, as will be seen, does not in other respects show much advance on Fig. 30.

Again in _Aphis_, the embryology of which has been so well worked out by Huxley,[18] the case is very similar, although the legs are somewhat later in making their appearance. When the young was 1/140th of an inch in length, he found the cephalic portion of the embryo beginning, he says, "to extend upwards again over the anterior face of the germ, so as to constitute its anterior and a small part of its superior wall. This portion is divided by a median fissure into two lobes, which play an important part in the development of the head, and will be termed the 'procephalic lobes.' I have already made use of this term for the corresponding parts in the embryos of Crustacea. The rudimentary thorax presents traces of a division into three segments; and the dorso-lateral margins of the cephalic blastoderm, behind the procephalic lobes, have a sinuous margin. It is in embryos between this and 1/100th of an inch in length, that the rudiments of the appendages make their appearance; and by the growth of the cephalic, thoracic, and abdominal blastoderm, curious changes are effected in the relative position of those regions."

In _Chrysopa oculata_, one of the Hemerobiidæ, Packard has described[19] and figured a stage in which the body segments have made their appearance, but in which he says "there are no indications of limbs. The primitive band is fully formed, the protozorites being distinctly marked, the transverse impressed lines indicating the primitive segments being distinct, and the median furrow easily discerned." Here also, again, the dorsal walls are incomplete, and the internal organs as yet unformed.

In certain Dragon-flies (_Calepteryx_), and _Hemiptera_ (_Hydrometra_), the legs, according to Brandt,[20] appear at a still earlier stage.

According to the observations of Kölliker,[21] it would appear that in the Coleopterous genus _Donacia_ the segments and appendages appear simultaneously.

Kölliker himself, however, frankly admits that "meæ de hoc insecto observationes satis sunt manca," and it is possible that he may never have met with an embryo in the state immediately preceding the appearance of the legs; especially as it appears from the observations of Kowalevski that in _Hydrophilus_ the appendages do not make their appearance until after the segments.[22]

On the whole, as far as we can judge from the observations as yet recorded, it seems that in Homomorphous insects the ventral wall is developed and divided into segments, before the appearance of the legs; but that the latter are formed almost simultaneously with the cephalic appendages, and before either the dorsal walls of the body or the internal organs.

As it is interesting, from this point of view, to compare the development of other Articulata with that of insects, I give a figure (Fig. 32), representing an early stage in the development of a spider (_Pholcus_) after Claparède,[23] who says, "C'est à ce moment qu'a lieu la formation des _protozonites_ ou segments primordiaux du corps de l'embryon. Le rudiment ventral s'épaissit suivant six zônes disposées transversalement entre le capuchon anal et le capuchon céphalique."

Among Centipedes the development of _Julus_ has been described by Newport.[24] The first period, from the deposition of the egg to the gradual bursting of the shell, and exposure of the embryo within it, which, however, remains for some time longer in connection with the shell, lasts for twenty-five days. The segments of the body, originally six in number, make their appearance on the twentieth day after the deposition of the egg, at which time there were no traces of legs. The larva, when it leaves the egg, is a soft, white, legless grub (Fig. 33), consisting of a head and seven segments, the head being somewhat firmer in texture than the rest of the body. It exhibits rudimentary antennæ, but the legs are still only represented by very slight papilliform processes on the undersides of the segments to which they belong.

As already mentioned, it is possible that at one time the vermiform state of the Homomorphous insects--which, as we have seen, is now so short, and passed through at so early a stage of development--was more important, more prolonged, and accompanied by a more complete condition of the internal organs. The compression, and even disappearance of those embryonal stages which are no longer adapted to the mode of life--which do not benefit the animal--is a phenomenon not without a parallel in other parts of the animal or even of the vegetable kingdom. Just as in language long compound words have a tendency to concision, and single letters sometimes linger on, indicating the history of a word, like the "l" in "alms," or the "b" in "debt," long after they have ceased to influence the sound; so in embryology useless stages, interesting as illustrations of past history, but without direct advantage under present conditions, are rapidly passed through, and even, as it would appear, in some cases altogether omitted.

For instance, among the Hydroida, in the great majority of cases, the egg produces a body more or less resembling the common _Hydra_ of our ponds, and known technically as the "trophosome," which develops into the well-known Medusæ or jelly-fishes. The group, however, for which Prof. Allman has proposed the term Monopsea,[25] and of which the genus _Ægina_ may be taken as the type, is, as he says, distinguished by the absence of a hydriform stage, "the ovum becoming developed through direct metamorphosis into a medusiform body, just as in the other orders it is developed into a hydriform body." Fig. 34 represents, after Allman, a colony of _Bougainvillea fruticosa_ of the natural size. It is a British species, which is found growing on buoys, floating timber, &c., and, says Allman,[26] "when in health and vigour, offers a spectacle unsurpassed in interest by any other species--every branchlet crowned by its graceful hydranth and budding with Medusæ in all stages of development (Fig. 35), some still in the condition of minute buds, in which no trace of the definite Medusa-form can yet be detected; others, in which the outlines of the Medusa can be distinctly traced within the transparent _ectothèque_ (external layer); others, again, just casting off this thin outer pellicle, and others completely freed from it, struggling with convulsive efforts to break loose from the colony, and finally launched forth in the full enjoyment of their freedom into the surrounding water. I know of no form in which so many of the characteristic features of a typical hydroid are more finely expressed than in this beautiful species."

Fig. 36 represents the Medusa form of this species, and the development thus described may be regarded as typical of the Hydroida; yet, as already mentioned, the Æginidæ do not present us with any stage corresponding to the fixed condition of _Bougainvillea_, but, on the contrary, are developed into Medusæ direct from the egg.

On the other hand, there are groups in which the Medusiform stage becomes less and less important.

The great majority of the higher Crustacea go through well-marked metamorphoses. Figs. 37 and 38 represent two stages in the development of the prawn. In the first (Fig. 37), representing the young animal as it quits the egg, the body is more or less oval and unsegmented; there is a median frontal eye, and three pairs of natatory feet, the first pair simple, while the two posterior are two-branched. Very similar larvæ occur in various other groups of Crustacea. They were at first regarded as mature forms, and O. F. Müller gave them the name of Nauplius. So also, the second or Zoëa form (Fig. 38) was at first supposed to be a mature animal, until its true nature was discovered by Vaughan Thompson.

The Zoëa form of larva differs from the perfect prawn or crab in the absence of the middle portion of the body and its appendages. The mandibles have no palpi, the maxillipeds or foot-jaws are used as feet, whereas in the mature form they serve as jaws. Branchiæ are either wanting or rudimentary, respiration being principally effected through the walls of the carapace. The abdomen and tail are destitute of articulate appendages. The development of Zoëa into the perfect animal has been well described by Mr. Spence Bate[27] in the case of the common crab (_Carcinus mænas_).

All crabs, as far as we know, with the exception of a species of land crab (_Gegarcinus_), described by Westwood, pass through a stage more or less resembling that shown in Fig. 38. On the other hand, the great group of Edriopthalma, comprising Amphipoda (shore-hoppers, &c.) and Isopoda (wood-lice, &c.) pass through no such metamorphosis; the development is direct, as in the Orthoptera. It is true that one species, _Tanais Dulongii_, though a typical Isopod in form and general character, is said to retain in some points, and especially in the mode of respiration, some peculiarities of the Zoëa type; but this is quite an exceptional case. In _Mysis_, says F. Müller,[28] "there is still a trace of the Nauplius stage; being transferred back to a period when it had not to provide for itself, the Nauplius has become degraded into a mere skin; in _Ligia_ this larva-skin has lost the traces of limbs, and in _Philoscia_ it is scarcely demonstrable."

The Echinodermata in most cases "go through a very well-marked metamorphosis, which often has more than one larval stage.... The mass of more or less differentiated sarcode, of which the larva, or pseud-embryo, as opposed to the Echinoderm within it, is made up, always carries upon its exterior certain bilaterally-arranged ciliated bands, by the action of which the whole organism is moved from place to place; and it may be strengthened by the super-addition to it of a framework of calcareous rods."[29] Müller considered that the mouth and pharynx of the larva were either absorbed or cast off with the calcareous rods, but were never converted into the corresponding organs of the perfect Echinoderm. According to A. Agassiz, however, this is not the case, but on the contrary "the whole larva and all its appendages are gradually drawn into the body, and appropriated."[30]

Fig. 39 represents the larva of a sea-egg (_Echino cidaris_) after Müller.[31] The body is transparent, shaped somewhat like a double easel, but with two long horns in front, which, as well as the posterior processes, are supported by calcareous rods. This larva swims by means of minute vibratile hairs, or ciliæ. It has a mouth, stomach, and in fact a well-defined alimentary canal; but no nerves or other internal organs have yet been discovered in it. After swimming about in this condition for a while, it begins to show signs of change. An involution of the integument takes place on one side of the back, and continues to deepen till it reaches a mass or store of what is called blastema, or the raw material of the animal body. This blastema then begins to change, and gradually assumes the form of the perfect Echinoderm.[32]

Fig. 40 represents a larva, probably of another sea-egg (_Echinus lividus_), from the Mediterranean, and shows the commencement of the sea-egg within the body of the larva. The capital letters denote the different arms: _a_ is the mouth, _a'_ the oesophagus, _b_ the stomach, _b'_ the intestine, _f_ the ciliated lobes or epaulets, _c_ the young sea-egg.

The development of the beautiful _Comatula rosacea_ (Fig. 41) has been described in the "Philosophical Transactions," by Prof. Wyville Thomson and Dr. Carpenter.[33] The larva quits the egg, as shown in Fig. 42, in the form of an oval body about 1/30 inch in length, something like a barrel, surrounded by four bands or hoops of long vibratile hairs or ciliæ. There is also a tuft of still longer hairs at the narrower posterior end of the body. Gradually a number of minute calcareous spines and plates make their appearance (Fig. 43) in the body of this larva, and at length arrange themselves in a definite order, so as to form a bent calcareous club or rod with an enlarged head.

As this process continues, the little creature gradually loses its power of swimming, and, sinking to the bottom, looses the bands of ciliæ, and attaches itself by its base to some stone or other solid substance, the knob of the club being free. The calcareous framework increases in size, and the expanded head forms itself into a cup, round which from five to fifteen delicate tentacles, as shown in Fig. 44, make their appearance.

In this stage the young animal resembles one of the stalked Crinoids, a family of Echinoderms very abundant in earlier geological periods, but which has almost disappeared, being, as we see, now represented by the young states of existing more advanced, free, species. This attached, plant-like condition of _Comatula_ was indeed at first supposed to be a mature form, and was named Pentacrinus; but we now know that it is only a stage in the development of _Comatula_. The so-called Pentacrinus increases considerably in size, and after various gradual changes, which time does not now permit me to describe, quits the stalk, and becomes a free _Comatula_.

The metamorphoses of the Starfishes are also very remarkable. Sars discovered, in the year 1835, a curious little creature about an inch in length, which he named _Bipinnaria asterigera_ (Figs. 45-47), and which he then supposed to be allied to the ciliograde Medusæ. Subsequent observations, however, made in 1844, suggested to him that it was the larva of a Starfish, and in 1847 MM. Koren and Danielssen satisfied themselves that this was the case.

Figs. 45 and 46 represent the front and side view of a Bipinnaria found by Müller[34] near Marseilles. _a_ is the mouth, _b_ the oesophagus, _c_ the stomach, _c_' the intestine. Fig. 47 represents a somewhat older specimen, in which the Starfish (_k_) is already beginning to make its appearance.

But while certain Starfishes thus go through metamorphoses similar in character, and not less remarkable than those of sea-eggs, there are others--as, for instance, the genus _Asteracanthion_--in which development may be said to be direct--the organs and appendages special to the Pseud-embryo being in abeyance; while in another genus, _Pteraster_, they are reduced to a mere investing membrane.[35]

Among the Ophiurans also we find two well-marked types of development. Some passing through metamorphoses, while others, as for instance _Ophiopholis bellis_, "is developed very much after the method of _Asteracanthion Mülleri_, without passing through the Plutean stage."[36]

Even in the same species of Echinoderm the degree of development attained by the larva differs to a certain extent according to the temperature, the supply of food, &c. Thus in _Comatula_, specimens which are liberally supplied with sea-water, and kept warm, hurry as it were through their early stages, and the free larva becomes distorted by the growing Pentacrinus (see Fig. 43), almost before it has attained its perfect form. On the other hand, under less favourable conditions, if the temperature is low and food less abundant, the early stages are prolonged, the larva is longer lived, and reaches a much higher degree of independent development. Similar differences occur in the development of other animals, as for instance, in the Hydroids,[37] and among the insects themselves, in Flies;[38] and it is obvious that these facts throw much light on the nature and origin of the metamorphoses of insects, which subject we shall now proceed to consider.