i8 9 5. NOTES AND COMMENTS. 219 



Are Small Variations Eliminated ? 



We have explained that Professor Weldon found that the 

 variations in the dimensions of the frontal length and of the right 

 dentary margin in the case of over 7,000 specimens of female shore- 

 crabs followed probability. His next step was to sort his specimens 

 into groups of approximately the same age, and to investigate the 

 occurrence of the variations at these ages. In both dimensions 

 measured, he found that there was a period of growth at which the 

 frequency of deviations increased. In fact, he had practical illustra- 

 tion of Darwin's statement that variations frequently appear late in 

 life. But there was an important difference between the two cases. 

 The preliminary increase in deviations was followed, in the case of 

 the frontal breadths, by a decrease in deviations of a particular 

 magnitude. In the case of the right dentary margins, no decrease 

 followed in later life. Assuming that abnormal young crabs tend to 

 grow up into abnormal older crabs, that, in fact, abnormalities do not 

 tend to correct themselves during growth, it is clear from the data 

 that there must be a special death-rate among crabs which have a 

 certain deviation from the normal in the case of the frontal breadth, 

 and that there is no special death-rate among crabs with deviations 

 of the right dentary margin. Experiments are in progress to test 

 whether or no the assumption that these abnormalities do not correct 

 themselves during growth be true. But, upon the assumption, it 

 appears that there is a selective destruction among crabs with 

 particular deviations from the normal frontal breadth. Professor 

 Weldon estimates this death-rate as amounting to seven per cent. 



It is to be noticed that the results obtained by this statistical 

 method recall other facts in very important fashion. They were 

 obtained from specimens of the shore-crab taken in Plymouth Sound. 

 In a communication made by Professor Weldon to the Royal Society 

 more than a year ago, he showed that among crabs collected at 

 Naples a marked dimorphism existed in this matter of frontal breadth, 

 and that variations of certain other organs were associated with 

 dimorphism of the frontal breadth. Moreover, frontal breadth is a 

 character which for long has been taken by zoologists as of systematic 

 value. The importance of these results seems to us more than a 

 reward of the enormous labour by which they were obtained. 



The Statistical Method in Variation. 



Professor Weldon's results have already established the impor- 

 tance of his method, and we cannot doubt but that wherever the 

 method be applied with equal diligence and discrimination, equally 

 important results will be attained. But we cannot agree that the 

 questions of species and of changes in them are to be dealt with 

 without taking function into account. It is clear that statistical 



s 2 



