124 The American Naturalist. [February, 
In order that we may compare, part with part, the different 
forms of Arthropods, it becomes necessary to assume some 
basis of comparison, and apparently the only one available is 
that of the exact homology of the similarly situated meta meres 
in the different groups, but here we meet with a difficulty. 
How can we be certain, for example, that somite 10 of the lob- 
ster is the exact homologue of somite 10 in the beetle? How 
can we tell that no somite has been lost in the evolution of 
these different lines? Perfect certainty is impossible, and we 
now know that in the serial comparisons of not more than five 
years ago, errors crept in, because there is a tendency of somites 
to become aborted or obsolete. This tendency is well-known 
in cases of Apus and Oniscus, where one of the anterior pairs 
of appendages is greatly reduced ; and in Limulus, Scorpions, 
Moina, etc., where an anterior somite is not differentiated until 
after those behind it. In many forms there is an obliteration 
or a fusion of cœlomic cavities in the anterior region, the 
mesoderm flowing together as a common mass. 
On the other hand, the embryonic phases of the nervous 
system seem to give clear indications of neuromeres in the 
anterior end of the body, and, as farther back, neuromeres cor- 
respond to the mesodermic metameres, it is reasonable to accept 
until error be shown, a somite for each neuromere at the an- 
terior end of the Arthropod body. Unfortunately, we have 
detailed knowledge of these neuromeres in but few cases, and 
even in these there is a lack of uniformity in the observations. 
In the Hexapods it has been shown that the “cerebrum” 0 
the adult is composed of at least three pairs of ganglia called 
by Vaillanes, respectively, the protocerebrum, the deutocere 
brum and the tritocerebrum. These elements have been rec 
ognized by Tichomiroff in the silkworm (teste Cholodkowsky) 
in Acilius (Patten, ’88), in Blatta (Cholodkowsky, ’91), m 
Mantis (Vaillanes, 91), in Xiphidium and Anurida (Wheeler, 
93), while Carrière (90) has described four cerebral elements 
in Chalicoderma. The figures of the latter author do not seem 
to me conclusive, and I am inclined to believe the more num- 
erous observations in this difficult field as the more probab 
correct 
