10 U.S. NATIONAL MUSEUM BULLETIN 255 



Measurements 



The wing expanse of all specimens was measured as the maximum 

 distance from the apex of one forewing to that of the opposite wing. 

 In certain instances the length of the forewing is indicated and was 

 measured with the aid of a Bausch and Lomb microscope sHde pro- 

 jector and a stage micrometer; an outline of the wing was projected 

 on to paper, marked, and then measured directly with the projected 

 image of the micrometer. A similar method was used to note measure- 

 ments of the male genitalia and, particularly in the case of the latter, 

 is believed to be quite accurate (within 0.01 mm.). 



A few common statistics have been included, particularly in the 

 discussion of Tegeticula yuccasella, and wlU be defined here. The 

 standard deviation (s) was calculated by the method given by Simpson 

 et al. (1960, pp. 84-87), which is based on the ratio: 



N 



^~ N-l 



where 



/= frequency of a given class 

 X= value of the variate for that class 

 A7^= number of observations (total frequency) 



The coefficient of variabifity (C.V.) is an expression of the standard 

 deviation as a percentage of the sample mean and is calculated as : 



X 



where 



x = sample mean 



The procedure followed in determining the statistical significance 

 between certain measurements treated in this paper is that of Snedecor 

 (1956, pp. 244-245, 250), whose test criterion, or F value, is calculated 

 as: 



J-, mean square of sample means 



mean square of individuals 



If the computed F value has been found to be significant at the .05 

 level of probabiHty, it is noted with a single asterisk (*); if its prob- 

 ability Hes beyond the .01 level (highly significant), then it is marked 

 with a double asterisk (**). 



