4 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.61. 



this nonconformity to type will be of most interest when it is possible 

 to compare it with similar studies made on parasitic insects in other 

 groups. This deviation is not so much to be regarded as a failure 

 of heredity or a wandering from a fixed standard as a normal 

 tendency of a species to vary, and of this inherent tendency biolo- 

 gists and systematists must take account. Some species may vary 

 more than others, and by the use of the statistical method it is 

 possible to derive and record considerable information concerning 

 a species, information of a very different character from that usually 

 given in descriptions and yet quite as fundamental as color, vestiture, 

 or sculpture. The purpose of systematic entomology is to classify 

 insects into groups of individuals known as species. But the present 

 tendency is to describe the type specimen (an individual) implying 

 that this description will fit other individuals of the group, then some- 

 times telling how the paratypes differ from the type and calling the 

 result a description of a species, without mention of any species 

 characteristics, such as range in size or any measures of prevailing 

 type or variability, qualities which belong to the group as a whole 

 and might serve to distinguish it from other related groups. The 

 biometrical constants and frequency distribution curve are constant 

 for the species and have the advantage of not being expressed in 

 terms of any individual in the group. 



In the accompanying graphs (fig. 1) the full line (for females) 

 and the long and short dotted line (for males) represents the original 

 data on the frequency distribution of the sizes of males and females 

 plotted as ordinates, the measurements being grouped into half- 

 millimeter classes and the abscissas taken from the mid-point of each 

 class. By mathematical methods these ordinates have been graduated 

 and the resulting smoothed curves represented by dotted (for 

 females) and crossed (for males) lines. The computed constants 

 are as follows: 



It will be seen that by any measure of prevailing type, the mode, 

 median or mean, the females are somewhat larger than the males. 

 The variability of the females is greater also. It is no greater, how- 

 ever, than that found in a gall-forming species ^ studied by the 



sCanad. Ent.. vol. ."1, 1010, pp. 



