1154 The American Naturalist. [December, 
Yet how can structures have the same phylogenetic origin when 
they, to all appearances, arise now from the ectoderm, now from 
the entoderm? The ingenious idea of Kleinenberg,. that “ the 
sexual cells do not come from the germ layers,” will help us out 
of this difficulty when properly applied. He says further “ that 
they already existed in the ancestors of the €celenterates when 
composed of loosely arranged similar cells, not yet differentiated 
into ectoderm and entoderm.” I would here replace “ Coelen- 
terate” by “ Metazoa,” since I cannot regard these radiate crea- 
tures as the ancestors of the Bilateralia, but only as animals in 
which the structure of the body has undergone this special 
transformation, owing to a previous sedentary mode of life. 
This, indeed, may well have been the case in all animals with 
radiate symmetry. 
Such primary germ-cells would then have originally formed 
the origin of the secondary or ccelomic mesoderm, and hence 
belong as little to one as to the other of the two primary germ: 
layers, but are merely interpolated for a time amongst the ele- 
ments of one layer or the other in the beginning of ontogeny in 
the Metazoa. Only we are, however, not able to distinguish them 
from their neighboring cells. So that it does not signify if they 
Subsequently move into the primary body-cavity as “ pole-cells,” 
or temporarily remaining in their first surroundings, furnish cell 
masses growing into the blastoccel, or else by multiplying in situ 
form epithelial surfaces that subsequently become completely 
Separated." Thus, as far as the coelomic mesoderm is concerned, 
the discussion as to its ectodermal or entodermal origin becomes 
quite unnecessary. Since the primary germ-cells probably lay 
on the boundary between the outer and inner layers, where they 
obtained both favorable conditions of nutrition and the possi- 
bility of discharging their derivatives by the shortest route, they 
could, later on, get into the ectoderm as easily as into the ento- 
rm. 
" By a similar method Rabl has recently shown how the coelomic diverticula of the 
"iod a ae modes of development of such structures ; but he is in error when 
oderm universally takes its origin from the entoderm. 
