COLORATION IN POUSTES. 15 



Moreover, the color areas being parts of a spherical surface and exceed- 

 ingly irregular, it was considered impracticable to measure them or to 

 make any attempt to seriate them in the methods usually employed 

 in statistical variation. The accompanying scheme, therefore, shows 

 only approximately the distribution of the material among the various 

 types, but it clearly shows what is the modal condition of the species, 

 as far as the dorsal surface of the abdomen is concerned, and what 

 are the various trends of development from this modal condition to 

 the more aberrant forms. 



The continuity of the variatioyi. — Considering classes c' and c, the 

 differences with regard to the abdomen are certainly great enough to 

 entitle the possessor to different specific rank, but when the gradual 

 transitions between the conditions are observed, the claim to specific 

 rank vanishes. The same is true of the fuscous coloring (PI. I, figs. 7 

 and 8). This reminds one strongly of P. canadensis, but the stages by 

 which it is reached are so easy and numerous that here again no claim 

 to specific rank could be maintained. 



As this study proceeds it becomes apparent that the three series of 

 transitions just described are in the direction of three different species, 

 namely, P. aurifer, a dark variety of P. pallipes, and P. canadensis; 

 but what is quite as apparent is that the differences between the ele- 

 ments of the series are exceedingly minute and insensible. We have 

 here absolutely no evidence for the necessary di.scontinuity upon which, 

 as a factor in the evolutionary process, Bateson lays such stress. The 

 processes by which the color is laid down are essentially chemical, and 

 it is just for these processes that Bateson claims and, it would seem, 

 claims justly, the necessity of discontinuity. But although the nature 

 of the chemical processes and the character of the various steps has 

 been, in a manner, ascertained, the final results of the process, at least 

 in this species, form continuous series, the exact position of whose 

 terms has been with difficulty determined, owing to the extreme degree 

 of imperceptibility in the differences involved. 



Degree of Variabii^ity in Males and Females. 



Taking the collection as a whole, it was desirable to learn whether 

 the males and females differed in the degree or kind of variability from 

 the modal condition. For this purpose the distribution of the two 

 sexes in the general scheme is plotted in fig. 8. The females are 

 represented by unshaded, the males by slightly shaded areas. From 

 this it is evident that the sexes agree in the manner of divergence 

 from the type pattern, but the females are characterized by much 



