840 The American Naturalist. [October, 
out detailing the very numerous and striking contrivances displayed by 
nearly all the anterior legs of some Craspedosomatids, it is sufficient to 
point out that in Scytonotus* modifications apparently as great as those 
of Glomeridesmus occur as far back as the twentieth pair of legs. In 
the light of these facts the degree of modification shown by Glomeri- 
desmus counts for little or nothing as an evidence of relationship with 
the Oniscomorpha. It might be said that Glomeridesmus has n9 cop- 
ulatory legs at all, for the structures figured by Mr. Pocock are prob- 
ably not comparable with the true copulatory legs of the other Diplo- 
pod groups, either in structure or function. The really remarkable 
thing about Glomeridesmus is that the legs of the seventh segment are 
not modified. Yet on this account we are not obliged to arrange 
Glomeridesmus in a separate category, for the degrees of modification 
to be found in the legs of the seventh segment of the other Diplopod 
groups are very various. It is even possible to trace, in the second pair 
of legs of the seventh segment of Craspedosomatidee all the stages from 
the nearly normal to the completely modified condition. Thus with 
reference to the fact that the seventh legs are unmodified, Glomerides- 
mus may be looked upon as one end of a series, not necessarily farther 
removed from the other groups than they are from each other. Cer- 
tainly the distance between the unmodified legs of Glomeridesmus and 
the distinctly jointed copulatory legs of Polydesmoidea and Polyzonoidea 
is not greater than that between those of the Polydesmoidea and the 
Spirostreptoidea. 
If, however, we admit that differences in the position and degree of 
modification of legs transformed for copulatory purposes are not of 
themselves characters of fundamental importance in the Diplopoda, we 
may seem to be under the necessity of admitting in addition that the 
constant appearance of what we may call the true copulatory legs in 
the place of the anterior or both pairs of the seventh segment is an evi- 
dence that the Helminthomorpha of Pocock are a homogeneous 
group to the extent of having more affinity with each other than 
with the Oniscomorpha. If, however, facts exist which indicate that 
the copulatory legs may have had independent origins in any of the 
Helminthomorpha the evidence just referred to is largely overthrown, 
for the utter diversity in plan of the copulatory legs of the different 
orders of Diplopoda is itself a strong indication that they represent 
independent lines of development. Such seems to be the import of the 
fact that the legs which in the Merocheta‘ are primarily transformed 
3 Am. N. Y. Acad. Sci., VIII, 233 (1894). 
t An ordinal name to cover the Polydesmoidea, Craspedosomatoidea and Calli- 
podoidea. Cf. Ann. N. Y. Acad. Sci., Vol. IX, (1895). 
