418 Mr. A. Alcock on 
mobranch fishes in which the female develops during 
pregnancy a vast system of uterine glands that secrete a 
nutrient fluid, or uterine milk, for the nurture of the deve- 
loping embryo. 
In this paper there will be given a detailed account of the 
phenomenon as lately reinvestigated in the species—Trygon 
Bleeker?, Blyth—in which it was first noticed by us. 
As is well known, reproduction among the Elasmobranchii 
is effected by the internal impregnation of the female. 
In some, as in the familiar instance of the ray, the female 
after impregnation lays eggs, which are enveloped in a tough 
leathery capsule secreted by the oviduct. 
In others, as familiarly exemplified by many sharks, the 
ege undergoes its changes and the embryo completes its 
development in the terminal part of the oviduct, which is 
now enlarged and elaborated to form a true uterus for the 
reception and retention of the embryo. In this case, as has 
long been known, a true placenta is formed, differing from 
the Mammalian placenta in the particulars which follow from 
the one main general fact that it is the yolk-sac, instead of 
an allantois, that furnishes the foetal part of the structure. 
There is yet a second method of viviparity known to occur 
among the Elasmobranchs, and to some particulars of it this 
paper is devoted. In this method while, on the one hand, 
the egg is retained and the embryo nourished within an 
oviductal enlargement or uterus, on the other hand no sort of 
vascular connexion is formed between the parent and the 
foetus. Here the expenditure of tissue comes altogether from 
the maternal side, the whole of the egg being devoted to the 
foetus and none of it being set aside to form vascular absorbent 
structures. 
In passing, one cannot but remark upon the interesting fact 
that in the primitive Hlasmobranch group we find in co-exist- 
ence all the methods of reproduction that occur in the higher 
Vertebrate phyla, namely (1) oviparity, with large-yolked eggs 
enclosed in a more or less rigid shell, (2) viviparity, with 
the formation of a placenta, and (3) aplacental viviparity. 
So far as we are at present aware the method of utero- 
gestation now under consideration reaches its perfection in 
the Batoidei; and of the six families into which this suborder 
is divided it has been observed in three, namely the Torpe- 
dinide, the Trygonide, and the Myliobatide. 
In Torpedo, as Professor Wood-Mason and I have elsewhere 
recalled, 1t was investigated in furthest detail by Dr. John 
Davy, who, in pregnant females, noticed (1) fcetuses lying 
naked in the uterus and unattached to it by any form of 
