EARLY ONTOGENETIC PHENOMENA IN MAMMALS. 149 
that phylogeny—about which paleontology will never be able 
to instruct us—on quite another basis. 
This basis is far from being complete. ‘Too little is yet 
known of the histological detail of the placentation process 
in the greater majority of mammals, and even when we will 
be fully acquainted with all those details as far as the recent 
mammals are concerned, even then we will perceive that the 
clue to many questions of phylogenetic importance lies 
amonest the extinct genera. 
We may, however, say, that if on the one hand placenta- 
tion details will help us to establish natural affinities in the 
grouping of the mammals, on the other hand no phylogeny 
of the placenta should be considered admissible if it would 
lead to any artificial grouping of naturally allied or naturally 
diverse mammals such as was discussed on p. 150, 
Viviparity and placentation have gone hand in hand with 
the development of allantois and amnion. And only after 
the two latter had appeared in the early viviparous tetrapods 
of the paleeozoic period did certain side lines of development 
diverge from that which led up to modern Mono- and 
Didelphia. 
In these side lines oviparity again came to the front, and 
on them we meet the parent forms of the Ornithodelphia, 
the Reptilia, and the birds. 
CHaptrer VI.—REFLEXIONS ON THE PHYLOGENY AND THE 
Systematic ARRANGEMENT OF VERTEBRATES. 
We have in the preceding chapters attempted to establish 
that certain fundamental conceptions concerning the em- 
bryonic envelopes and the placentation of the higher verte- 
brates are much in want of a renewed critical analysis. We 
have some time ago (’02, 705) come to a similar conclusion 
with respect to gastrulation in vertebrates.' 
1 Keibel (705) and Brachet (05) have expressed their conviction that these 
