
ACROLOPHUS. 319 
genera, and have since described two more genera as new—these have been generally 
accepted and maintained, after further study, by Fernald, Beutenmiiller, Druce, and 
Dyar. With a very large accession of material before us it became necessary closely 
to examine all doubtful cases and to verify the special points which had been relied 
upon in distinguishing the numerous genera established in what has been regarded as 
a fairly well-defined family. We approached the subject with a very keen desire not 
only to maintain these divisions but, if possible, to find within the group such 
characters as would justify their further sub-division, and it was only after a prolonged 
struggle on these lines that we were obliged to admit that there could not be found 
any character, or combination of characters, on which these genera, or any one of them, 
could be maintained in such manner as to be made recognisable either by figure or 
description. The intermediate variations were so numerous and so graduai that, apart 
from the question of the value of secondary sexual characters, which has been con- 
sistently discounted in this paper and the uselessness of which was very specially 
illustrated in this particular group, we were driven to seek for differences in neuration 
after all other means of distinguishing the genera had been found unreliable. Here, 
as in other respects, the disappointment was great. We found that these Acrolophids 
were in such an evidently plastic condition that even in the same species, and in some 
cases on the two sides of the same specimen, the apical veins, whether 7 and 8, or 8 
and 9, were either forked, connate, or separate, apparently without any controlling 
rule or fixity of purpose. We were somewhat prepared to find the apical veins variable 
in this respect, and it was not until the position of the internal veins of the cell 
(medial and radial) had been carefully studied, in relation to the point of origin of the 
apical veins running to the costa and termen, that we were forced to abandon any 
attempt to define the particular series to which veins 7 and 8 should be assigned. 
This was found to be fairly consistent up to a certain point, and for some time we felt 
justified in retaining the genus Caenogenes, upon the ground that 7 and 8 were stalked 
from below the internal radial, but even this distinction ultimately broke down, and 
in one species forrert we found that vein 8 was sometimes stalked with 7, sometimes 
stalked with 9, sometimes separate from both, and at other times connected with either 
in varying degrees, moreover 7+8 originated sometimes from below and sometimes 
from above the internal radial. In one specimen of another species, cleptica ( 3 66895), 
not only 7 and 8, but 9 and 10 also were distinctly stalked. 
Therefore, much against our united inclination and efforts, it has been necessary 
to regard the so-called Acrolophidae (with two exceptions) as one large genus, and to 
rely upon such differences, as had hitherto been regarded of generic value, for the 
purpose only of separating the very numerous species within it—the form of the genital 
segments will be found at least as useful for that purpose as any other character. 
After dismissing the attempt to found any separate genera on the radial neuration 
of the forewings, it may yet be doubtful whether the cubital veins do not afford reliable 
3 ¢¢ 2 
