This page needs to be proofread.
974
EMBRYOLOGY


which fed on the debris of both animal and vegetable nature in the undergrowth of primeval forests. Such is in fact the life of the lowest insects known to-day, some of which, such as Machilis, nourish them- selves on the decaying sea-weed on the sea-shore, and retain through- put life vestigial limbs attached to the abdominal segments which aid them in their crawling movements. Now the myriapod or poly- pod larva survives as the caterpillar of the Lepidoptera and the primitive Hymenoptera. It is also found amongst the primitive Meuroptera and amongst the may-flies (Ephemeroplera). These last-named insects were supposed to possess a larva showing great secondary modifications, for it is provided with leaf-like gills at- tached to its abdominal segments; but Heymons 1 has shown that these gills are nothing but modified abdominal legs. It is a curious fact that amongst the lower insects, such as the cockroach, this polypod stage is passed through during the embryonic phase of development. The reason for this change is that these insects lay their eggs in situations where a grub-like larva would perish, whereas the higher insects, in which the stage is larval, are gifted with in- stincts which lead them to lay their eggs in situations where an abundance of easily procured and easily masticated food is available and a scavenging existence like that of the ancestor is possible.

A fourth factor which modifies development, and which is potent in its effect although it is usually overlooked, is loss of size in the larva as compared with the ancestor which it represents. As conditions change and the larval life becomes more dangerous there arises a tendency in which we may trace the influence of tachygenesis to pass quickly through the larval stage and to metamorphose at as early a period of growth as possible into the adult condition. A consequence of this change is that the larva assumes a new relation to its environment, for many qualities of the surrounding medium, such as the viscosity and support ing- power of water, acquire an altered importance as the organism decreases in size. If the an- cestral organs were reduced in the same scale as is the whole body of the larva, this would in many cases result in their becoming in- capable of being used. As a consequence we find that In many cases where the ancestor had a series of organs, this series is represented in the larva by fewer members or only one member of larger relative size, and that where in the ancestor there was a pair of organs there is frequently only one in the larva, but this is on a larger scale than the rest of the body. If we now select a few examples to illustrate this principle, we may consider the free-swimming larva of that most primitive of all vertebrates Amphioxus. This larva has only a single series of gill-slits which are so enlarged as to occupy the whole ventral surface of the body. If the double series of slits, which the ancestor in common with all other fish possessed, had been developed in the larva, they would necessarily have been of such minute size that the capillarity of the water would have prevented them from being functional. Similarly there is no doubt that the eyes of verte- brates were from the beginning paired structures, but they are repre- sented in the ascidian tadpole by a cup-like outgrowth of one side of the brain. Again no serious doubt can be entertained that the primitive arthropod was evolved from a long many-segmented annelid with flexible parapodia. But the most primitive larval form of the Crustacea is the nauplius which is a little, oval, unsegmented creature with but three pairs of legs. The existence of this larva was a great stumbling-block to the earlier embryologists. It seemed to indicate that the Crustacea must have been derived from an unsegmented animal totally distinct from the ancestor of other Arthropoda, for the progenitor of these must have been long and segmented since the embryos of all these Arthropoda have many segments. But if we look at the nauplius larva from the standpoint of function rather than of structure we have no difficulty in seeing in it the recapitulation of the first step in the " arthropodization " of the annelid. This step was a change of habits which consisted in using the foremost parapodia as oars to propel the animal and as organs to seize food. As a consequence in the front of the body the cuticle was thickened and the " arthropodous " type of limb pro- duced, whilst in the rest of the body the annelid condition of affairs persisted as indeed it may be almost said to do in the posterior portions of the bodies of this most primitive of Crustacea, the long- bodied Phyllopoda such as Artemia. Once begun in front, this " arthropodous " modification was gradually propagated backwards so as to involve the hinder segments of the body and in this way the higher Crustacea were evolved. In the nauplius larva, the anterior arthropodized portion of the ancestor with its appendages is alone represented; the hinder annelid portion in which function was less intense and less important is not developed.

One last instance of the principle may be adduced which we select from the embryology of the higher vertebrates. In the development of those types of Vertebrata in the life history of which there is a long larval phase (Cyclostomata, dipnoan, "ganoid" and teleostean fish, Amphibia), a larval excretory organ termed the pronephros makes its appearance. Its duct later becomes the duct of the perma- nent kidney, but the pronephros itself consists of very few tubules and these originate from the wall of the general body cavity and not, as do the tubules of the permanent kidney, from special sacs

1 R. Heymons, " Cber die Lebensweise und Entwicklung von Ephemera vulgata," Sitzungsberichte der Gesellschaft der Natur- forschenden Freunde zu Berlin for 1896.

(the malpighian capsules). The earlier workers regarded the pro- nephros as a last trace of a primeval excretory organ of quite different structure to the permanent kidney by which it is later superseded. The works of Hatta 2 on the development of the lamprey and of Kerr 3 on the development of the Polypterus have proved that the pronephros is nothing more than the foremost section of the perma- nent kidney, early called into action and enlarged whilst the hinder section of the metamerically repeated series of tubes of which the kidney consists remains undeveloped. These investigators have fur- ther proved that the portion of the general body cavity from which the pronephric tubules arise consists of several malpighian capsules fused together and secondarily communicating with the general body cavity.

Life History of Animals. It might be supposed that with so many modifying factors at work it would be a hopeless task to attempt to disentangle the recapitulatory element from them, and that therefore the ancestral history of animals except in its latest and least modified chapters would remain a closed book. But when we recollect that the life history of every species con- stitutes a separate edition of this history, and that the modifying factors have affected no two of them to the same extent, it be- comes evident that comparative embryology built on a broad basis can attack the problem with a fair prospect of success.

Bearing in mind the priority of the larval over the embryonic phase, and beginning therefore our survey with the larvae of the simplest metazoa, we are able to recognize the first step in the evolution of the metazob'n from the protozoa in the blastula, the hollow ball of cells which may be regarded as representing a colonial protozoon like the living Vohox. This stage was followed by the formation of a gut by the intucking of one side of the ball; and this second step is represented in the life histories of all the lower and simpler animals by the gastrula stage. Following on this stage came the formation of the coelom as a series of pouch- like outgrowths of the gut, and the change of the single opening of the gastrula, the blastopore, into two openings which became the mouth and the anus by the constriction of its middle portion.

It has been possible to show that two groups so utterly diverse in appearance as the Annelida and the Mollusca have originated from a single group of free-swimming ancestors represented by the trochophore larva, and since Arthropoda are admitted by all to be descended from Annelida this conclusion involves the ancestry of four-fifths of the animal kingdom.

We can form a very plausible guess as to the nature of the diver- gence of habits which led to the differentiation of the Annelida and Mollusca from one another. The original stock was free-swimming but both groups derived from it are typically bottom-dwellers. Two modes of seeking their food were open to them ; they could either glide over the bottom by means of their cilia as young Gastropoda and Lamelli branchiata still do, or they could burrow into it. The first led to the evolution of Mollusca, the second to that of Annelida.

Two other groups of very diverse structure, which embryology has given strong reasons for believing to have been derived from a single race, are the Echinodermata and the Vertebrata. The lowest form which gives distinct evidence of the vertebrate affinities is the worm-like creature Balanoglossus. The larval form of Balanoglossus is a free-swimming organism called tornaria which shows the closest resemblance to the typical larva of Echinodern'ata, the dipkwvla. The recognition of this affinity has assisted in the elucidation of a difficult subject to which considerable space was devoted in the llth ed., viz. the origin of the central nervous system.

It is characteristic of the most primitive Annelida and Arthropoda that this system develops as a ring round the blastopore and an endeavour was made in the nth ed. to prove that this was originally true of vertebrate embryos also. But it is a peculiarity of the verte- brate-echinoderm alliance which is still unexplained that in them the blastopore gives rise to the anus alone, whilst the rronth is formed as an apparently independent perforation at a considerably later period of development. A long succession of embryologists, with their eyes fixed only on the embryos of Annelida, Arthropoda and the higher vertebrates, have held that the vertebrate month is a new structure formed by the fusion of a pair of gill-slits and have prosecuted vain searches for traces of the old mouth. Others have imagined that the mid-dorsal line of the vertebrate embryo along which the nerve-cord develops corresponds to the line joining mouth and anus in the arthropod, the line in fact which is occupied by the slit-like blastopore in Peripatus. They hold that the nervous system of the vertebrate originally extended round -the front end of

2 Hatta, " The Development of the Renal Organs in the Lam- prey," Journ. Coll. Sci. Imp. Univ. Tokio, 1912.

3 Kerr, " Vertebrata with the exception of Mammals,' Textbook of Embryology, vol. i., pp. 223-237 (1919).