300 



Embryogenesis: Progressive Differentiation 



material necessary for gastrulation upward 

 from the equator, reaching the blastodisc 

 at about the time of the third cleavage in 

 Carassius but later in Fundulus. 



The types of abnormality appearing in 

 cultured blastodiscs seem not to be the same 

 in Carassius as in Fundulus, and may imply 

 some morphodynamic divergence. In the 

 former, trunk axis seems to differentiate in 

 preference to head or tail; in the latter, 

 head differentiates preferentially whereas 

 trunk and tail are most easily suppressed. 

 Devillers ('49) suggests that these differences 

 may be referred to distinctive patterns of 

 axial elongation in the two species. This 

 is probably only part of the story; interpre- 

 tation should perhaps be deferred until the 

 cotirse of morphogenetic movements in 

 Carassius is understood. 



In addition to the latitudinal separation of 

 the cleaving egg, Tung and Tung ('44) have 

 succeeded in a remarkable feat, the meridi- 

 onal bisection of the whole egg at the 

 2-celled stage or later, followed by the rear- 

 ing of both halves. Nine such pairs are de- 

 scribed in their report: Figure 110 attempts 

 to summarize these descriptions. It is seen 

 that the successful pairs range from two 

 perfect embryos down to two non-gastrulat- 

 ing fragments. Some particularly interesting 

 pairs occurred: one case of two complete 

 embryos showed slight complementary asym- 

 metry reminiscent of amphibian embryo 

 pairs obtained by constriction methods 

 (Spemann and Falkenberg, '19). In another 

 remarkable pair, the notochord is confined 

 to one member, and is disproportionately 

 large; the nerve cord appears qviite normal 

 in the deficient twin, where the position of 

 the chorda is occupied by median somites. 

 The interpretations that unilateral localiza- 

 tion (in the yolk mass) of material neces- 

 sary for gastrulation is responsible for this 

 series of results, or that some spatial localiza- 

 tion of chorda material in the blastodisc 

 is indicated, are possible but not necessary 

 conclusions from such a series as this. 

 Clearly, defective gastrulation is indicated 

 in many cases; localized materials may or 

 may not play a role. 



In svimmary, then, while some experi- 

 ments on Fundulus indicate great plasticity 

 of organization during cleavage, this form 

 is possibly not completely equipotent, nor, 

 probably, is Carassius. We are not yet in 

 a position to analyze these differences. In 

 both forms, some progressive stabilization 

 of material within the blastodisc seems to 

 occur, making it able to carry on develop- 



ment independently of the yolk over the 

 subsequent period. Devillers ('47, '49) has 

 shown that at the end of the cleavage 

 period, the blastodisc of the small-yolked 

 egg of Esox is similarly independent, 

 whereas that of Salmo must have a car- 

 bohydrate source in the cultvu'e medium. 



At present the cleavage stage in the 

 bird's egg is terra incognita experimentally; 

 it is possible only to make conjectures based 

 on the situation found after laying; this 

 will be discussed in the following section. 



EVENTS DURING GASTRULATION 



In the meroblastic eggs under considera- 

 tion, the blastula stage is the period during 

 which a cell size characteristic of the species 

 is being established in the blastodisc, which 

 thus acquires stable epithelial properties. A 

 constant subgerminal cavity indicates equili- 

 bration of fluid relations between yolk and 

 blastoderm. At this time are initiated the 

 major changes of form which continue 

 through gastrulation to the molding of the 

 embryonic axis. 



In the teleost, rapid spreading of the blas- 

 todisc is combined with thickening and in- 

 vagination of material at the peripheral 

 blastopore or germ ring. The hypothesis ex- 

 pressed by Lewis ('49) that this epiboly is 

 due to contraction of the yolk gel layer, 

 pulling the blastoderm down over the yolk, 

 has inspired considerable experimental work. 

 In the trout egg, Devillers ('51b) points out 

 that independent movements of the blasto- 

 disc in relation to its specific substratum 

 (the periblast) are of at least as much im- 

 portance as any gel-layer contraction. 

 Trinkaus ('51) gives strong evidence against 

 the effectiveness of any gel-layer traction in 

 Fundulus, and substitutes the interpretation 

 that the gel-layer controls and coordinates 

 growth of the blastodisc, but that the growth 

 itself is due to factors intrinsic in the disc, 

 or in its relation to the periblast. The latter 

 structure he has shown to have remarkable 

 independent powers: in the absence of a 

 blastodisc it can spread over the yolk and 

 even close a "blastopore." The gel layer, 

 he submits, solates as the periblast margin 

 advances, and hence cannot exert traction 

 on either periblast or blastodisc. 



In the avian egg, the blastoderm also 

 spreads rapidly over the yolk, and Schle- 

 singer ('52) has recently shown this to be 

 due to active growth of the cellular margin. 

 Movements of invagination, however, are 

 confined to the central posterior area, where 



