EARLY ONTOGENETIC PHENOMENA IN MAMMALS, 97 
On the other hand an embryonic shield, if curved in this 
way and then developing an early ventral mesoblast by 
proliferating ectoderm at the hinder end of the embryonic 
shield (cf. p. 92), would be very favourably situated for an 
early vascularisation of the trophoblast, all the more so if a 
decidua reflexa—as is also met with in many Rodents— 
co-operated. And so it would not seem impossible to turn 
the argument the other way, and to pretend that the very 
inversion was in some way connected with an earlier presence 
of a vascular “connective stalk.’ I do not, however, wish 
to press this argument, as I feel the ground is as yet too 
uncertain. I have only wished to call attention to those 
rodents and eventual insectivores where the first rudiment 
of the allantois is in no way an outgrowth from the hind wall 
of the gut, but simply a proliferation in a very early stage of 
vasifactive tissue at the hinder end of the embryonic shield. 
I may call attention to the fact that very many years ago [| 
have already suggested (’89, p. 375) that potentially the 
possibility of the development of a connective stalk was 
present in the hedgehog, and was only perhaps retarded by 
the fact of the omphaloidean placentation having acquired 
considerable significance in early stages. 
Summarising we may say that the allantois, such as we 
find it in Sauropsida and in many mammals and which has 
erroneously been looked upon as being primarily a urine- 
reservoir during embryonic life, is not the fit starting-point 
for phylogenetic speculations about its evolution, but that 
the more subordinate position in which we find it in the 
Primates offers a clue for the explanation of its earliest 
appearance. At the same time it makes us understand so 
much better, that the favourable conditions under which the 
Primates succeed in establishing an early and very thorough 
vascularisation of their trophoblast, has contributed largely 
to their exceptional development in regard to the central 
nervous system. 
We can safely say, as I wrote recently (’07, p. 60), that 
the different orders of mammals represent so many attempts 
VOL. 35, PART 1.—NEW SERIES. 7 
