78 A. A. W. HUBRECHT. 
123, 182). These latter figures, compared to Figs. 50, 57, 
68, 73—76, 82, 85 on Duval’s Pls. 2 and 3, make it clear to 
us how a case of closed amnion formation, as it is offered by 
Galeopithecus and Pteropus (and as it is virtually present in 
the very early bat stages (Duval’s Figs. 50 and 57), can gradu- 
ally become converted (in the other bats) in one in which the 
closed amnion only comes into definite existence through the 
gradual uprising of a rim, the outer wall of which is tropho- 
blastic, the inner one a derivate of the embryonic ectoderm. 
Many figures (8a, 13—17, 20, 23, 30—32, 37) have shown 
that the early separation between trophoblast and embryonic 
knob takes place in the most various ways. And _ that 
even in one and the same species, as, for example, 
Tarsius, the separation may come about earlier or later 
(cf. Hubrecht, ’02, Pl. II, fig. 38, a—e; with Pl. VI, figs. 
49, a, b, and 50 a—c). This later appearance calls forth 
a stronger resemblance between T'arsius and such cases as 
Pteropus or Cavia. At all events, it is this very early process 
of separation of what will be the embryonic tissues from the 
trophoblast that goes parallel to ontogenetic processes in the 
invertebrates to which we have called attention (Pilidium, 
Desor larva) as showing us the earliest causes of ammnio- 
genesis. Such cases as Tupaja (Fig. 30) and Cervus 
(Keibel, ’99), and lately again Sus (Fig. 17) are particularly 
instructive. What Selenka has designated by the name of 
Entypie is—from our point of view—no secondary pheno- 
menon, but one which repeats very primitive features of 
separation between embryonic ectoderm and larval envelope 
in invertebrate ancestors. 
The formation of a proamnion is a phenomenon which has 
no significance at all for our considerations concerning the 
phylogeny of the amnion in general. It is a temporary 
1 T call particular attention to Duval’s (99) Figs. 96 and 102, and feel too 
late, while correcting the proof, that I ought to have copied them in this 
paper, particularly because the independent growth of the trophoblastic 
(outer) and of the ectodermal (inner) plait of the amnion-fold is there so 
extraordinarily clear. 
