becomes divided by a constriction into mouth and anus, and that a portion of the gut epithelium, viz. that forming the midventral portion, is formed from ectoderm turned in round the edges of the slit. It is practically certain that this last statement also rests on an error of observation. In the primitive annelid Polygordius, VVoltereck ' has described a similar slit-like blastopore and he has followed the process of its closure in great detail describing the division of every cell involved. In this case the midventral epithelium of the gut is formed by the union of endoderm cells lying at the sides of the blasto- pore whilst the ectoderm cells lying in the blastopore lips by their union reconstitute the midventral skin. No reasonable doubt can be entertained that a renewed investigation with a more modern tech- nique would show that this is also true of Peripatus.
It must never be forgotten that embryological research is based on a comparison of embryos of different ages with one another not, as would be the ideal method, on a continuous observation of the progress of one and the same embryo. It follows that too large an age-difference between the embryos examined may give rise to a totally wrong conception of the process which is taking place. So are to be explained the statements which crop up from time to time, such as those of Heymons 2 that the mid-gut of the higher in- sects is entirely formed from ectoderm, and of Watase 3 who made a similar assertion about the mid-gut of the cephalopod Loligo. Hirsch- ler 4 has shown how the error of Heymons originated, and Watase has been corrected by Faussek 6 ; and should further statements of this kind occur in the literature the strong presumption is that they also are founded on mistakes.
Organ-forming Substances. We have arrived at the conclusion that the establishment of the validity of the germ-layer theory is one of the great achievements of embryological research, and we now turn to the question of how the differences which distinguish the layers from one another are brought about. We have learnt that in primitive alecithal eggs like those of Echinodermata all portions of the blastula wall are alike in their potencies and that the differentiation of ectoderm from endoderm only begins when the first traces of gut-formation are visible. We have likewise learnt that all parts of the primitive gut or archenteron are alike in their powers, and that the separation of endoderm from mesoderm only becomes apparent when the first indication of the coelom appears. But this progressive differentiation of the embryo might be due to a differentiation of the nuclei of various regions or of the cytoplasm or of both. We have, however, learnt from the development of the polyspermic frog's egg that there is a strong presumption that the nuclei of the embryo are alike in their nature and that the differentiation of the layers must be due to the separation of organ-forming cytoplasmic substances from one another. This conclusion is confirmed by a large number of observations on many different kinds of eggs; a few of the more striking may be given here.
Hertwig allowed frogs' eggs to develop under pressure between glass plates and in capillary tubes. Under these circumstances the divisions took place by planes normal to the pressure and flat plates and rows of cells were produced. When the pressure was removed, however, these deformed embryos recovered, multiplication of cells took place and the normal form was regained and normal develop- ment proceeded. It was easy to show that nuclei which under undisturbed conditions would have occupied certain definite regions of the embryo had been forced into quite other regions, and yet per- fectly normal embryos resulted. Hertwig concludes that the nuclei could be juggled about like a handful of balls without affecting the formation of the embryo.
In many eggs the differentiation of the layers is indicated at a far earlier period than that at which it occurs in the eggs of the Echinodermata or even of the lower Vertebrata like the frog. The egg of the ascidian Cynthia partita which has been studied in great detail by Conklin 7 may be adduced as an example. This egg when it develops becomes converted into an elongated blastula consisting
1 Woltereck, " Beitrag zur praktischen Analyse der Polygordius- entwicklung," Archiv f. Entwicklungsmechanik, vol. xviii., 1903.
2 Heymons, " Uber die Bildung der Keimblatter bei den In- secten," Sitzungsb. der Preussischen Akad. der Wiss., vol. i., 1894.
- Watase, " Observations on the Development of Cephalopods,"
Studies from the Biol. Lab. Johns Hopkins Univ., Baltimore., vol. vi., 1888.
- Hirschler, " Die Embryonalentwicklung y. Donacia crassipes,"
Zeitschrift fur ivissenschafUiche Zool., vol. xcii., 1909.
' Faussek, " Untersuchungen uber die Entwicklung der Cepha- lopoden," Mitteilungen a. d. Zool. Station zu Neapel, vol. xiv., 1920.
6 Hertwig, " Ueber den Werth der ersten Furchungszellen fur die Organbildung des Embryos," Archiv f. mikroscopische Anatomic, voK xlii., 1893.
7 Conklin, The Orientation and Cell-lineage of the Ascidian Egg," Journ. Acad. Sciences, Philadelphia. Series 2., vol. xiii., 1905.
of few cells; this blastula changes into a gastrula in the typical way, and though no distinct coelomic pouches are formed large portions of the archenteric wall are directly converted into muscles which lie at the sides of the tail of the tadpole-like larva. In this species the nucleus of the unripe egg is as usual a vesicle filled with fluid (the so-called germinal vesicle). The cytoplasm contains numerous yolk globules of a slaty-blue colour and also larger yellow- ish globules which are concentrated in its superficial layer. When the maturation divisions of the nucleus occur the nuclear wall is dis- solved and the fluid contents escape and form a cap of clear material at the animal pole of the egg. When fertilization takes place pro- found rearrangements of the substances in the cytoplasm are effected. The yellow globules stream downwards to meet the spermatozoon which enters at the vegetative pole, and they finally form a cres- centic layer of yellow material round the lower pole of the egg. As the egg develops first into a blastula and then into a gastrula, and finally into the characteristic ascidian tadpole, it becomes evident that the clear substance forms the ectoderm, the slaty-blue material the endoderm, whilst the yellow material forms the masses of meso- derm which give rise to the tail muscles. When the egg is in the four- cell stage the yellow material is confined to the two posterior cells; if one of these be killed the remainder of the egg will give rise to a larva with muscles only on one side of the tail. That the nuclei have nothing to do with this separation of substances is shown by what occurs at the lip of the blastopore. Here we find an arc of what Conklin calls " neurochordal " cells. Each of these has of course a single nucleus, but the cytoplasm of each consists of two zones, one clear and one slaty-blue. At the next division two daughter cells are produced from each neurochordal cell ; one of these contains the clear substance and is added to the nerve plate which is a part of the ectoderm; the other is composed of the blue substance and forms part of the notochord which in Cynthia as in other Vertebrata is a derivative of the endoderm.
From this development we conclude that the germinal layers owe their origin to the segregation of cytoplasmic substances in the growing e &&! that these substances assume their final arrangement under the influence of the spermatozoon, which thus on its path to meet the female pronucleus determines the symmetry of the embryo. Brachet 8 has shown that this is also true of the frog's egg. It was for long a puzzle why competent observers like Roux 9 and Hertwig 10 should differ so profoundly on the results of killing one of the first two blasto- meres of the frcg's egg. Roux asserted that the surviving blasto- mere gave rise to a half blastula which developed into a half tadpole, whilst Hertwig maintained that it tended to form a normal tadpole, being only impeded in its development by the mass of dead material constituted by the other blastpmere. Brachet has shown that both are right, for the plane separating the first two blastomeres need not by any means coincide with the future median plane of the embryo, but may make any angle up to a right angle with it. If it coincides with this plane by killing one blastomere Roux's result is obtained; if it is oblique the result accords with Hertwig's researches.
Thus the potency of each of the first two blastomeres of the frog's egg depends entirely on the cytoplasm it happens to include and in no way on the nucleus. Brachet " has shown that the fixing of the median plane of symmetry in the frog's tadpole, as in the ascidian tadpole, is effected by the spermatozoon. As the spermatozoon penetrates the egg in its path towards the female pronucleus, it leaves behind a trail of pigment which persists for a considerable time and can be detected at a much later period in the development of the egg. It is found that on the opposite surface of the egg to that at which the spermatozoon enters it, there is formed the so-called " grey crescent." This is in reality the upper lip of the blastopore; it is here that the differentiation of ectoderm from endoderm begins. Therefore we conclude that the arrangement of the organ-forming substances in the frog's egg is caused by the spermatozoon.
In the mollusc Dentalium when the egg has reached the four-cell stage one of the blastomeres emits a protuberance termed the " yolk-lobe " or " polar lobe." This lobe is devoid of a nucleus and before the attainment of the eight-cell stage is reabsorbed into the blastomere. Nevertheless, if this lobe be cut off, the remainder of the egg develops into a larva which is fatally devoid of mesoderm.
That the materials which form the basis of the different substances embodied in the germinal layers are formed in the growing egg under the influence of emissions from the nucleus is rendered certain first by the close relationship of the nucleus to assimilation and secondly by the fact (see CYTOLOGY) that the nucleolus of the unripe egg breaks up into fragments and is extruded into the cytoplasm. It is, however, a surprising fact that the nuclei of the segmenting egg are alike and apparently without influence on the differentiation of the primary organs. In fertilization a second nucleus of alien origin is introduced and portions of this nucleus, as we have already seen, are
8 Brachet, loc. cit.
- Roux, " Uber das entwicklungsmechanische Vermogen jeder
der beiden ersten Furchungszellen des Eies," Verhandlungen der anatomischen Gesellschaft, 1892.
10 Hertwig, " Uber den Werth der ersten Furchungszellen fur Organbildung des Embryos," Archiv fur mikroscopische Anatomic, vol. xlii., 1893.
11 Brachet, loc. cit.
