NORTH AMERICAN NOCTUIDAE— SMITH AND DTAR. 31 



of the ornainentation whatever changes might take place in other 

 directions. 



The organs that are most likely to be first affected are those of repro- 

 duction, and in the Noctuidae at least those of the male seem particularly 

 sensitive. I am not ready to claim that this is a rule, because our 

 knowledge of the life history of our species and of the larval structures 

 is too incomplete, but it is certainly true in Acronycta. A good species 

 is certain to liave specific peculiarities in all its stages, and if we do 

 not discover them the fault is with us. Any sound system of classifi- 

 cation based on one stage will harmonize with that built upon any other, 

 provided that the facts are rightly interpreted. If we accept evolution 

 as a fact it simply can not be otherwise, since all changes must have 

 acted upon the species as a whole, and larval characters could not be 

 continued unless the resulting adult was in turn affected. 



We find, therefore, that if we examine the genitalia of the male in 

 Acronycta^ the proposed classifications of Chai)man and Dyar, based 

 upon ljupae and larvae, are fully confirqjed, and that even more divisions 

 are indicated by this character. Dr. Chapman finds two very distinct 

 pupal types. Dr. Dyar finds four series in the larvae, of which two are 

 not sharply defined. In the adults there are five distinct genitalic types 

 and a sixth that, while recognizable, leaves a few intermediate species 

 unplaced. Unfortunately, too few of the larvae are known to make our 

 classification correspond completely, but there is no reasonable doubt 

 that our further knowledge will result in the discovery of characters 

 that will associate all those forms agreeing in genitalic structure. It is 

 interesting to note here that while excellent group characters are obtain- 

 able from these sexual structures, they are not of much specific impor- 

 tance in this genus, because all the species of each group resemble each 

 other quite closely in this particular. The group of which loheliac is 

 typical varies most, and it is from this that modifications seem to radiate. 



Five groups in this genus may be typified by americana, lobeliae, per- 

 suasa, hamamcHs, and auricoma, respectively. 



Group amcricand is well characterized by having, in the male, very 

 broad harpes or sidepieces, from almost the middle of which there arises 

 a long, curved, chitinous process. This i)rocess seems to arise directly 

 out of the membranous structure, and is not set on a chitinous ridge 

 or other separate strengthening process from the base. Superficially 

 the species agree in trigonate primaries, in which the outer margin is 

 long and evenly curved to the somewhat prominent apex. The macu- 

 lation, while complete in some cases, tends Co become broken up into 

 blotches, and the jw/ marks are rarely prominent. The i)alpi are shorter 

 than in any other group and hardly reach the middle of the front, which 

 in the species of this series tends very strongly to become conical and 

 somewhat prominent. The thoracic vestiture is decidedly less scaly 

 tlian in the other groups, and in some cases consists almost entirely of 

 long, very narrow, flattened hair. Altogether this group is fairly well 

 marked. 



