314 THE AMERICAN NATURALIST [Vol. XLIV 



is to be observed, then the individuals developed parthenogenet- 

 ically. This neat, though somewhat far-reaching, generalization, 

 rests solely and simply on the location by inspection alone of 

 modes or maxima of frequency in samples of from about 50 to 

 150 individuals! Nowhere in Bachmetjew's papers is there any 

 indication of such a concept as that the observed frequencies 

 might have a probable error ever having disturbed the placid 

 progress of the reasoning. 



To give these comments point some of Bachmetjew's frequency 

 distributions may be examined. Here are some distributions 

 supposed (with a mathematical naivete probably nowhere to be 

 paralleled in scientific literature) to show two maxima of fre-, 

 quency at the points indicated by the bold faced type 

 (Bachmetjew (5)). 



It certainly demands the most consummate scientific insight 

 to discern why the two modes in A and C, for example, should 

 be located as they are. And again the unenlightened statistician 

 wonders what it is about the second 15 in D which makes it a 

 mode rather than the first 15. Is it possessed of a different 

 essence or aura, say a sort of transcendental ' ' fif teenness " ? 



Or again, to take but one more example, who would have sup- 

 posed that distribution E had two modes while F had but one? 

 (Bachmetjew (5)). 



