FROM LAYING TO FORMATION OF FIRST SOMITE 85 
end. <A similar marginal notch at the hinder end of the blasto- 
derm in the line of prolongation of the primitive streak has been 
described also by His and Rauber, but in the cases observed 
by them there was no connection with the primitive groove. 
It suggested to them, however, the idea of genetic connection 
between the two, and was used as argument for the derivation 
of the primitive streak from the margin by concrescence. 
The second question concerning the primitive streak, its rdéle 
in the formation of the embryo, may be answered very briefly 
by saying that it is itself the primordium of the greater portion 
of the axis of the embryo; some indeed maintain that it represents 
the entire embryonic axis excepting the short pre-chordal part 
(Kkopsch). The view of Balfour and Dursy that it takes no essen- 
tial part in the formation of the embryo, but atrophies as the 
embryo forms, is now of historical interest only. The question 
is how much of the embryo is represented by the primitive streak. 
But this question is by no means easy to answer, and there is 
no complete agreement in regard to it. The one point that is 
definitely settled is that the anus arises at the hinder end of the 
primitive streak; but what point in the embryo corresponds to 
the anterior end of the primitive streak, or, in other words, how 
much of the embryo is laid down in the blastoderm in front of 
the primitive streak, is a disputed question. The attempt has 
been made to solve the problem by destroying the anterior end 
of the primitive streak by a hot needle, or by electrolysis, then 
sealing up the egg and permitting it to develop farther and finally 
locating the resultant injury in the embryo. But, while one 
worker finds the injury at the anterior end of the notochord 
(Kopsch), that is in the region of the fore-brain, another finds it 
in the region of the heart, that is in the hind-brain (Peebles). 
The reasons for this discrepancy in results are two: (1) the methods 
employed are not sufficiently exact, and (2) it is difficult in the 
living egg to determine the exact location of the anterior end of 
the primitive streak, and sometimes even to distinguish it from 
the head-process. Owing to the extremely rapid growth of all 
parts of the embryonie axis, a minute division of the primitive 
streak becomes a relatively long part of the embryonic axis in a 
very short time. It is obvious, therefore, that the slightest 
deviation of the injury from the point aimed at may lead to 
considerable error in the results. Until embryologists operate 
