118 NATURAL SCIENCE [August 



as a standard, then the same amount of variation in man and woman 

 has more effectiveness in one sex than the other, I must reply, no 

 one has yet investigated this point ; my own conclusions on skull 

 measurements, so far as they have yet gone, seem to show that the 

 co-efficient is at least a rough measurement of effectiveness. But it 

 must be clear that until we have investigated the relation of effective- 

 ness to some clear measure of variation, Darwin's law of the greater 

 variability of the male is entirely unproven. Whether we put 

 effectiveness as a function of mean and of standard deviation, or as 

 a function of mean and the ratio of standard deviation to mean, is 

 not, at first sight, a matter of great importance ; it is to be settled 

 rather by what the algebraist considers a convenient shape for his 

 formulae. It is the biologist who has to determine the form of the 

 function. It probably varies widely from species to species and 

 organ to organ, but it may reasonably be supposed to vary only 

 continuously and gradually with age and sex. If the selective 

 death-rate of any species, however, be a function only of the mean 

 and standard-deviation of any particular organ, then the theory of 

 dimensions shows us at once that the death-rate cannot be a 

 function solely of absolute variation, but must be a function of the 

 ratio of absolute variation to the mean, i.e., of the co-efficient of 

 variation. 



Lastly, if Professor Weldon thinks I have reviewed my biological 

 critic harshly, I would remark that I submitted my paper a year 

 ago in proof to a valued biological friend, I still have in a familiar 

 hand-writing "no suggestions to make." That Professor Weldon 

 should find in my paper a " violent assertion " to be regretted, con- 

 firms my view that modern biology is a house divided against itself. 



Karl Pearson. 



