HARLY ONTOGENETIC PHENOMENA IN MAMMALS. 79 
structure met with in many mammals and sauropsids where a 
circular region of the blastocyst in front of the head remains 
without mesoblast. Into this the front end of the embryonic 
body curves down and is temporarily sheltered. This 
envelope thus consists of ectoderm and entoderm only (Fig. 
150). It is during further growth of the embryo gradually 
reduced ; the head is withdrawn from it ; mesoderm gradually 
appears between its layers, and when the embryo is ready it 
has entirely lost its proamniotic covering layer which is 
finally flattened out... Thus the explanation of the amnion 
which appears furthest from the truth is that of Selenka 
(91, p. 186), who has expressed the opinion that the amnion 
arose out of a double source, and that the two Anlagen of both 
amnion and proamnion were finally fused into one. The true 
interpretation must decidedly start from quite different con- 
siderations as were developed above. 
1 The explanation of the first phylogenetic origin of a proamnion has not 
yet been attempted; generally it has been ascribed to rapidity of growth, 
which caused the embryo’s head to become temporarily imbedded in its own 
umbilical vesicle. But then the Primates, who have undoubtedly the biggest 
heads—comparatively !—have no proamnion. 
My own explanation is a very different one, and starts, not from yolk-laden 
eges of Sauropsids, but from early viviparous protetrapods which must have 
preceded (see p. 15) both Sauropsids and Mammals. Some of these have 
obtained direct vascularisation of the trophoblast by umbilical vessels; a 
great many others have departed from this very direct line of perfect vascula- 
risation, have given up the early ‘‘ Haftstiel” (the homologue of which 
reappears in the allantoidean attachment), and have at an early stage utilised 
their area vasculosa on the umbilical vesicle for establishing intercourse with the 
mother. ‘Thence arose a temporary omphaloidean placentation. After a time, 
however, the disadvantages of this system of vascularisation during the 
further increase in size of the embryo became evident. Not, however, before 
arrangements had come into existence by which the omphaloidean placentation 
could remain in function as long as possible. Of these arrangements the 
most important is no doubt the growth of tbe head down into the umbilical 
vesicle, with, as result, the formation of a proamnion, It reaches its maximum 
in the Didelphia which, after having given up allantoidean placentation (yet 
persistent in Perameles), enjoy omphaloidean attachment for a short time, and 
then come into the world under the quite specialised conditions that are so 
characteristic for the Marsupials. 
