270 
BULLETIN OF THE UNITED STATES FISH COMMISSION. 
course reject, because it is equivalent to admitting the homology of the blastoderm 
edge in Teleosts and Amniotes, and consequently the correctness of the Balfour-Eanber 
hypothesis; secondly, that the process is, to refer it to simple embryonic forms, one of 
progressive delamination. It will be seen that Kupffer’s hypothesis really implies the 
occurrence of the latter jirocess, for when he explains the spreading of the blastoderm 
as the completion of the blastula stage, he really means that the yolk splits off ecto- 
derm progressively from a and a' towards m. Thus, again to reduce the processes to 
their simplest forms, over one-half (yolk-half) of the blastula (Fig. 12, p. 269) delami- 
nation occurs ; but in the other half there is a true invagination (region of prostoma 
and primitive streak). IJeither Kupffer nor Hertwig illustrates his theory with dia- 
grams, and since the embryonic processes dealt with are extremely complicated, it is 
a difficult matter to form a precise conception of their meaning. However, I think 
the analysis I have given is a perfectly fair one, and the result is evidently prejudicial 
to their theory. For the conclusion is that the Amniotic vertebrates have a blastula, 
which invaginates over one half and delaminates over the other. Such an embryonic 
form is nowhere known to occur, and the theory which is forced to assume its exist- 
ence is in so far a weak theory and must give place to any other which can explain 
the facts by making use only of known processes. 
The explanation which Cunningham gives to the blastoderm edge, and to which 
Eabl is logically forced, is much simpler than the hypothesis just discussed. It is 
equally objectionable, however, for it invokes a process never observed, and which is 
moreover a priori extremely improbable. Adopting as an ancestral form a gastrula 
like that of tbe frog, Cunningham supposes the Amniotic gastrula to have been derived 
as follows : ^ The increase in size of the mass of yolk cells, instead of enlarging the 
blastopore (the result which one would suppose would naturally follow) produced a 
rupture in the ectoderm. In this way the lower surface of the yolk came to be ex- 
posed and the blastoderm edge was brought into existence, the latter being a purely 
secondary structure, dating no further back than the occurrence of the hernia. 
The fact that, having once accepted KupffeFs homology between the primitive 
streak and the blastopore we are forced to assume the occurrence of such improbable 
embryonic forms and processes as have just been described, leads us to reconsider the 
older Balfour Eauber hypothesis. The only objection to regarding the blastoderm 
edge as part of the original blastopore, is that round the latter there is an ingrowth of 
cells, round the former there is none. But the study of the meroblastic Teleost egg 
has already shown us that the part of the ingrowth which, according to the Balfour- 
Eauber theory, belongs to the blastoderm edge in the Amniota, became a rudimentary 
organ in the fishes. There is no more natural supposition than that it went a step 
farther and was altogether lost, or became still more reduced in size in the Amniota. 
The disappearance of a rudimentary organ is one of the commonest inferences in the 
study of the comparative anatomy of adults, and of course must occur in embryos as 
well. It is therefore a very different assumption from that of the occurrence of an 
embryonic hernia or a new and complicated embryonic form. Assuming that the meso- 
dermic ingrowth has been lost round the blastoderm edge in Amniota, the only objec- 
tion to the Balfour-Eauber theory falls to the ground, for of course this theory as well 
as the other is capable of explaining the existence of what Hertwig calls “unpaired 
mesoblast ” behind the primitive streak. 
It must be remembered, moreover, that the exact nature of the blastoderm edge in 
