STUDIES IN MAMMALIAN EMBRYOLOGY. 389 
genera have only been incidentally touched upon in this 
memoir, and it is not my intention here to work out this 
comparison more fully, which I reserve for a future publication. 
But I may be allowed to refer to these as yet unpublished 
observations as indicative that the subdivision of Deciduata 
will have to be carefully scrutinised, and that such scrutiny 
may lead to dissolution of the subdivision and to a total 
revision of what Huxley has called the “ placental ” classifica- 
tion, bringing it, to cite Huxley’s words (1. c., p. 101), “upon a 
more secure basis, if we take not merely with Milne-Edwards 
the form of the placenta, but with von Ber and Eschricht its 
structure into account.” Investigation of ‘‘ structure ” cannot 
nowadays neglect histological detail and ontogenetical data. 
If, then, I propose henceforth to abandon the subdivsion of 
monodelphous mammals into Deciduata and Adeciduata, 
I wish to support this proposal by yet another argument. This 
argument concerns the name selected by Huxley. I am con- 
vinced that it has already led to grave misunderstandings and 
to obscurity, both of expression and of representation. From 
the very first the name has been tainted by this congenital 
defect. In a footnote (1. c., p. 103) Huxley says: “It is of 
course by no means intended to suggest by these terms (the 
term: deciduate and non-deciduate) that the homologue of the 
decidua does not exist in the non-deciduate mammals.” This is 
filaments, which are often brought away, as in the rat, distinct from the 
maternal structure, like the foetal cotyledon in the cow.” In a footnote 
Owen adds: ‘‘ This never happens in the quadrumanous placenta; and the 
difference is not affected by showing that prior to severance the foetal placenta 
is combined with maternal vessels in the rat, mole, &c.” This was again 
noticed by Ercolani, and carefully inquired into. In his summary Ercolani 
says (‘Memorie Accad. Scienze Ist. Bologna,’ vol. x, p. 912) that in the 
mole: “non esiste alcun intimo rapporto fra la porzione fetale e la materna 
della placenta che si snucleano assai facilmente fra di loro...... lo 
snucleamento delle dette parti avvenga in questo animale con una facilita 
assai piu grande di quella che deve avvenire nei cotiledoni di alcuni cervi.” 
These points cannot, however, be further discussed here, but will tend to 
awaken a healthy scepticism with respect to the rigorously deciduate character 
of the Insectivora in particular, and thus to the primary significance of the 
deciduate subdivision in general. 
