STUDIES IN MAMMALIAN EMBRYOLOGY. 307 
ture of trophoblast and allantois. The trophoblastic blood- 
spaces, enclosed as they are in thin layers of cellular tissue, 
show a marked tendency to bulge out into the cavity (the extra- 
embryonic ceelom) in which the allantois is developing. In 
many of the preparations this roof of diplotrophoblast, against 
which the allantois is going to extend, instead of being flat, as 
fig. 53, is mammillated. The surface is undoubtedly increased 
in that way, and the allantois finds the surface against which it 
is going to be applied as it were prepared for the reception of 
allantoidean outgrowths, ridges and villi. That this is in fact 
the way in which the allantois finally effectuates its contact 
with the trophoblast, is demonstrated by numerous preparations 
in my possession. One of these is represented in fig. 54. It 
is there, moreover, seen that the somatic mesoblast can also 
in this mammillated phase of the trophoblast be sometimes 
detected as a very thin membrane between trophoblast and 
allantois, which here and there would even seem to be inter- 
rupted, and which at all events does not in its turn proliferate 
as do both the trophoblastic and the allantoidean tissues.!. The 
further growth and development of these two tissues leads 
to the formation of the definite allantoidean placenta, simulta- 
neously with the phenomena above described of the retro- 
gressive metamorphosis aud final disappearance of the ompha- 
loidean placentation. The allantois is at the outset a thick- 
walled sac with very numerous blood-vessels in its walls, and 
witli strands of tissue radially connecting these walls at different 
points. It retains these general characteristics during its 
further growth, the walls becoming relatively thinner as the 
size of the allantois increases, and the radial strands forming 
pillar-like communications between the opposite surfaces. 
Blood-vessels of the allantois, leaving the body of the embryo, 
1 For Vespertilio murinus Frommel describes and figures (I. c., pl. 
vi, fig. 13) a preparation which is in many respects comparable to what 
I here described for the hedgehog. A certain amount of proliferation of the 
somatic mesoblast (though not very conspicuous) appears to have been 
noticed by him in this mammal, where the mammilliform projections of the 
trophoblast offer undoubted analogy to what I have noted for the identical 
region in Hrinaceus, 
