58 U.S. NATIONAL MUSEUM BULLETIN 265 



previously, some correlation may exist between the relative sizes of 

 the male and female genitalia; however, for another character studied, 

 that of relative wing width, similar correlation does not seem to be 

 present. Certain specimens from Aiizona (Madera Canyon) and 

 California (Lucerne Valley) consistently demonstrated narrow wings, 

 similar to the moths from Torrance Co., New Mexico, but differed 

 considerably from the latter population in possessing male genitalia 

 of a much shorter length. After several more populations have been 

 surveyed and additional characters studied, such discrepancies may 

 become clarified. 



The 1222 specimens examined during the course of this study 

 represented a major portion of T. yuccasella's known distribution. 

 Following a casual examination of this series, it soon became evident 

 for both sexes that the eastern forms generally possessed broader 

 wings than the western ones (figs. 86-87). As a means of presenting 

 this variation, 14 localities, consisting of a total of 193 females, were 

 selected to represent the species range (see table 3) . These particular 

 localities were chosen because it is believed that most, if not all, 

 constitute definite populations. That is to say, all specimens (for a 

 given locality) were collected at the same site, during the same time 

 period by the same collector. For some locales, host data was also 

 available to associate the material further. The forewings of all 

 examples were measured directly with a stage micrometer by pro- 

 jecting an enlarged outline of the wing onto paper. It is believed 

 that wing measurements performed in this manner are accurate to 

 within Koth of a millimeter. To summarize these measurements, the 

 average width-length ratios were calculated (WXIOO/L) for each 

 locality; the standard deviation (s) and coefiicient of variability 

 (C.V.) of the ratios are included in table 3 for comparisons. Most 

 values of C.V. were between 4 and 6.8; some localities exhibited a 

 lower value, reflecting in most instances a small sample size. All 

 data listed in table 3 are based on measiu-ements of female specimens; 

 however, in order to check for possible sexual differences between 

 males and females in regard to wing width, some males were likewise 

 examined. For example, 13 males from Madera Canyon agreed 

 closely with their corresponding females in possessing a mean wing 

 ratio of 28.18. 



The general geographic distribution of relative wing width appears 

 more definite than that of genitalia length; although, again, no clear 

 lines of demarcation can be observed. Wing width appears to decrease 

 gradually from east to west, with some overlap in the Southwest. 

 Further collecting may show that the distribution of this character is 

 somewhat similar to that of male genitalia size (i.e., in direct relation 



