54 NATURAL SCIENCE [July 



No one acquainted with the facts of animal variation — certainly 

 not Professor Pearson himself — will assert that the " coefficient of 

 variation " is always a measure of the importance of variations ; no 

 one will believe that in any animal a deviation of ten per cent, in 

 excess of the mean of one organ has of necessity the same im- 

 portance as a deviation of ten per cent, in excess of the mean in 

 another organ. A finger nail of double the normal length, or a hair 

 of double the normal thickness, will hardly produce so much incon- 

 venience as a leg of double the normal length. It is even certain 

 that in closely allied species the same per centage deviation of 

 corresponding characters may produce widely different effects. 



There is no doubt that the " coefficient of variation " is for 

 certain purposes a valuable measure of variability ; and Professor 

 Pearson has shown, in some of his more technical papers, that it is 

 of great use in establishing important propositions in the theory of 

 Chance. At the same time, students of evolution, paying attention 

 specially to the functional importance of variation, may need units 

 proportional to this importance ; and such units may well be different 

 in different cases. The violent assertion that there is only one 

 " scientific " measure of variability is therefore to be regretted. 



Whether one agrees with Professor Pearson on this single question 

 or not, one cannot but be grateful to him for the four essays here 

 referred to, as well as for the more elaborate memoirs on which they 

 are based. The picture they present of the orderly treatment to 

 which animal statistics can be subjected, so that hitherto unwieldy 

 and perplexing masses of figures can be made to yield simple and 

 intelligible results, should do much to make the study of Probability, 

 in its application to the problems of animal evolution, more popular 

 than it is, and to enable biologists at last to put before themselves an 

 adequate numerical estimate of those phenomena which it is the 

 business of their lives to formulate and to explain. 



W. F. R. Weldon. 



