66 



U.S. NATIONAL MUSEUM BULLETIN 2 75 



other abnormalities were not used. Series of specimens were analyzed 

 by standard statistical methods which involved computation of the 

 arithmetic mean, standard deviation, and standard error of the mean. 



In some instances, in which the absolute size of a character is the 

 same in two taxa, scatter diagrams are employed to demonstrate 

 differences in the relative size of this character. 



Graphical methods employed here are patterned after those given 

 by Hubbs and Hubbs (1953). The range of variation is shown by a 

 heavy horizontal line, and the mean is indicated by a sharply pointed 

 triangle. 



The black portion of each bar comprises two standard errors of the 

 mean on either side of the mean. The entire bar (black and white 

 portions) constitutes one standard deviation on either side of the 

 mean. In samples of four specimens, two standard errors and one 

 standard deviation have the same value, and the entire bar is there- 

 fore black. Samples of less than four specimens were not treated 

 statistically. In these cases, the range of variation is shown by a 

 heavy horizontal line, the mean is indicated by a sharply pointed 

 triangle, and the bar is omitted. A single triangle indicates that only 

 one specimen is represented. The standard deviation (S.D.) indicates 



Figure 3. — Principal collecting localities. 



