272 THE DOCTRINE OF DESCENT, 
mammals, the vascular processes and vilii of the wall 
of the ovary become so closely connected with the foetal 
portion of the placenta, that at birth the entire mem- 
branous coating of the ovary is detached and thrown out 
with it. In others, the vascular villi are not so closely 
adherent; they yield without important lacerations, and 
hence no deciduous membrane (Membrana decidua) is 
ejected. Now, asit appears to me, the specific conditions 
of the formation of the decidua have been far too little 
compared to justify our inferring any close affinity from 
the mere fact that portions of the coating of the ovary 
are lost in parturition, Much rather it must be unre- 
servedly admitted that the formation of decidua might 
be occasioned by subordinate circumstances of the most 
varied kinds, and hence in orders only remotely allied, 
or allied merely as placental mammals. We therefore 
consider the decidua to be a subordinate systematic 
feature where anatomical and morphological reasons ~ 
are opposed to it. 
We go yet further. In the modern system the form 
of the placenta js likewise employed in the grouping of 
organisms. If among the Deciduata, lemurs, rodents, 
insectivora, bats and monkeys, are classed together as 
orders with discoidal placenta, this combination is cer- 
tainly supported by a series of other reasons, and it is 
quite probable that within this group of orders the form 
of the placenta is due to homology, that is to Descent. 
But when beasts of prey, elephants and the Daman 
(Hyrax) are further cited as orders with zonary pla- 
centa, we find ourselves in the same position as when 
the decidua was reckoned decisive as to the. closer 
affinity; and we are of opinion that the subordinate form 
