224 
to be atypical enough throughout to suggest extensive deviations in 
details. 
While my work has been done principally on the Arthropods, I 
have examined representatives of other phyla, and in a tentative 
study of the mouse testis have clearly distinguished a structure which 
so closely resembles the accessory chromosome of the insect cells that 
I am convinced of its identity. 
From the results so far obtained, I was assured that we had 
to do with a structure of great importance and one which would 
repay most careful study. I have therefore devoted my attention 
almost exclusively to it, and together with my students, have collected 
over a thousand testes from nearly a hundred species for study. It 
is only by an extensive comparative study, as I suggested in my first 
paper, that any reliable conclusion may be hoped for. 
The theoretical portion of the large paper was written with much 
reluctance. The small amount of undisputed information at hand 
made generalization a difficult matter: but a working hypothesis is 
necessary, and in view of the fact that others have been published, 
I decided to include mine with the observations. Before presenting 
this, however, the view advanced by PAULMIER, and adopted with 
more or less reservation by WıLson and MONTGOMERY, may well be 
considered. 
Because the accessory chromosome fails to divide in the second 
spermatocyte, PAULMIER considers that it is a chromosome in the 
process of disappearing from the species. As evidence against this 
assumption, I would point out that the history of the element in the 
spermatogonia, where it is raised to the rank of a nucleus by se- 
parate inclusion in its own vesicle and where it normally divides in 
each mitosis; its regular and constant behavior in the first spermato- 
cyte, where it is remarkable for its unvarying position and staining 
reaction; and finally the uniformly undivided condition itself in one 
of the spermatocytes all show that the element is a normal one. 
Degenerate structures are always irregular and uncertain in their 
manifestations, while in the case of the accessory chromosome we 
have the greatest degree of regularity and certainty of behavior. 
PAuLMIER’s theory, however, breaks down completely when the 
true character of the accessory chromosome is considered. Nearly all 
observers are agreed that the element is a spermatogonial chromosome 
which passes over entire into the spermatocytes. The appearances in 
the Orthoptera are, I think, conclusive proof of this. It will be noted 
further that the element retains its form as a chromosome unvaryingly 
