304 A. A. W. HUBRECHT. 
grows larger, the hinder part of the trunk and tail are turned 
about 90° (cf. diagram, fig. 33), so that the embryo turns one 
side of its trunk and tail towards the allantoidean trophoblast, 
the other towards the yolk-sac. The head all the while pro- 
jecting into the cavity of the yolk-sac. With the further 
increase of the yolk-sac the head is retracted from its pro- 
amniotic covering, and the whole embryo has then assumed a 
position in which one of the sides of both head and trunk are 
turned towards the allantoidean trophoblast. The diagram, fig. 
34, corresponds with this stage. During the later stages of preg- 
nancy the embryo again turns over to its original position, the 
middle of the back finally facing the middle of the placenta 
(diagram, fig. 36). This turning process is simultaneous with 
the arrest in growth of the yolk-sac, and it is thereby that the 
latter is so neatly folded. Different stages of the folding are 
in my possession as separate preparations, but more space need 
not here be devoted to a full description of them. 
In the later stages of pregnancy we have seen that the om- 
phaloidean trophoblast has played out its part, that it becomes 
thinner and thinner, and finally membranaceous, together with 
what we will have to describe as the decidua reflexa in the 
next chapter. We know that the villi of the yolk-sac are then 
retracted from their corresponding crypts, and we have only 
still to inquire whether any special phenomena can yet be 
noticed concerning the thin layer of somatic mesoblast which 
was already noticed in the incipient stage of the yolk-sac de- 
velopment, which is figured in fig. 43. It follows, from my 
preparations, that this layer does assume a peculiar aspect. 
When a larger and larger portion of the surface of the yolk-sac 
is loosened from its close connection with the trophoblast the 
extra-embryonic ccelom increases, and the somatic and splanchnic 
portion of the mesoblast are simultaneously definitely separated. 
The thin layer of somatic mesoblast, which I have already 
referred to above, and which in its relation to the allantois 
will be discussed by-and-by, undergoes a very distinct modifi- 
cation in those regions where the yolk-sac was formerly 
adherent to the trophoblast, i.e. all along that portion of the 
