1897.] On the Affinities of Tarsius : 573 
of the origin of the placental “ anlage” in Tarsius the whole blas- 
tocyst is lined by the mesoblast and in the portion of the latter 
which connects the placental anlage with the posterior end of 
the embryo, originates a very important structure, highly char- 
acteristic of the Anthropoids; this is the bauchstiel or ventral 
stalk. Prof. Hubrecht explains the genesis of the bauchstiel 
very clearly, and shows that as the embryo becomes folded 
off from the yolk by the development of the amnion, that this 
mesoblastic strip is carried with the embryo until the latter is 
really suspended in the cavity of the blastocyst, its only attach- 
ment to the wall of the same is by means of the bauch- or 
haftstiel as he calls it. Vascularization of the ventral stalk 
soon takes place by blood-vessels arising in it, and thus em- 
bryo and placenta are brought into vascular connection. 
Another very important point shown by Hubrecht is that the 
portion of the bauchsteil next to the embryo is permeated by 
two tubes, the dorsal is a prolongation of the amnion and the 
lower is the rudiment of the allantois. The latter is one of 
the most important discoveries made by Hubrecht in connec- 
. tion with the development of Tarsius, as he demonstrates that 
in Tarsius, as in the Apes and Man, the allantois is rudimen- 
tary and does not hang freely in the coelomic cavity as in the 
rest of the Mammalia. Prof. Hubrecht shows that the allan- 
tois rudiment takes no share in the vascularization of the 
bauchstiel as the following quotation proves:-“ Finden wir in 
dem Haftstiel eine verhaltnissmiissig langen Restbestand dies- 
es Allantois rohres, welches aber, wie bereits auseinander gesetzt 
wurde, an der Vaskularisation des Haftstiels keinerlei antheil 
hat.” Professor Hubrecht compares an early stage of Erinace- 
us, before the splitting of the mesoblast (see his paper, stage 
2), with that of Tarsius (fig. 7), in order to prove that Erinace- 
us has a sort of bauchstiel, connecting the rear end of the 
embryo with the trophoblast. In this stage of Erinaceus the 
extent of the mesoblast is very limited, but in Tarsius the 
whole blastodermic vesicle is lined by mesoblast. 
Turning now to the foetal membranes of the Lemurs, we find 
there is apparently little in common between the placentation 
of these animals and Tarsius. In Tarsius the placenta is dis- 
