EARLY ONTOGENETIC PHENOMENA IN MAMMALS. 113 
production of certain distinct cellular (decidual) elements, 
which might approach the denuded surface and eventually 
fuse with or pass through the adherent trophoblast cells. 
The maternal capillaria would then not either be eaten into, 
but also have preserved their endothelium, consequently the 
interaction between the maternal and the embryonic blood 
would be a little less direct, but this might be balanced by this 
somewhat less direct interaction taking place over a consider- 
ably extended surface, consequent upon the so much more 
considerable size of the blastocyst. Now if we consult the 
very recent paper by Ciro Barbieri (’06) on the placentation of 
Tragulus meminna we finda state of things there described 
which approaches closely to what was sketched above, viz. a 
denuded mucosa, an active trophoblast of which vascularised 
villi penetrate into denuded crypts, decidual maternal elements 
which pass from the mucosa into the trophoblast, thus form- 
ing an association of maternal and embryonic elements as in 
the dog, not, however, localised, but transitory. 
The fact that the surface of the interchange between the 
Tragulus foetus and its mother has a more considerable 
surface extent than that in the dog and very much more than 
in any Insectivore should also not be lost sight of, especially 
when we consider that other species of Tragulus examined 
by Selenka (91) and by Strahl (05) show, again, further 
diminution of the destructive trophoblastic activity, because 
in these the maternal epithelium does not disappear in the 
erypts. ‘The maternal and embryonic blood is in these cases 
separated by fully two epithelial strata, and the passage of 
decidual elements through the trophoblast was not noticed. 
There is, of course, not the least difficulty in passing from these 
latter stages on to those which we find both in Ruminants 
and in such Ungulates as the horse and the pig, which latter 
have always been looked upon as the prototype of the diffuse 
placenta. 
Suppose this to have been the real phylogenetic develop- 
ment of the arrangement of the so-called ‘ placenta” of 
Ungulates—which would thus in reality be a secondarily 
8 
