22 NATURAL SCIENCE. January, 



species, but also the number of individuals exhibiting any particular 

 variation. We can thus see at a glance the proportion between the 

 average and the departures from it, and if we should find, by com- 

 paring curves constructed from the examination of individuals of 

 different ages, that the death-rates of the possessors of different 

 variations were different, we should have reasonable ground for 

 supposing that natural selection was weeding out certain variations. 

 We should further have an answer to the question whether the species 

 was being modified or not — this answer being positive or negative 

 according as the death-rate was greater on one side of the average 

 than on the other, or equal on both sides. 



The main objections which can be urged against this method are 

 of a practical nature. First, we isolate a character, and try to 

 determine whether or no its possessor suffers by its presence. But 

 " characters " are mere mental conceptions, they do not exist by 

 themselves, and natural selection acts, so to speak, on the balance of 

 all the characters. Of course, if we could prove that individuals 

 possessing a certain character had a lower death-rate than individuals 

 without it, we might expect that that character would become a 

 distinguishing feature of the species, though the survival of those that 

 possessed it might not be due to its utility, but to some constitutional 

 peculiarity of which it was the by-product. How are we to determine 

 such an association of death-rate and variation ? The attempt has 

 been made by drawing a curve for the same character in the case of 

 young and of older animals. The curve was flatter in the case of the 

 younger animals ; that is, the number of deviations from the mean was 

 found to be greater in them : hence it is assumed that these abnormal 

 forms are continually being weeded out. It is possible, however, that 

 there may be a self-regulating tendency in growth. Are not 

 abnormalities observed in children often toned down in later years ? 



There is another objection, the validity of which would not be 

 admitted by many, but which seems to me to have a certain weight, 

 and that is this : So far as we can discover, the condition of the organic 

 world has remained relatively stable during the historical epoch, and 

 no new species are known to have been formed. It is a question, 

 therefore, on the one hand whether the conditions which brought 

 about the great variability of which we see evidence in the geological 

 record, still persist, and if not, whether the process of evolution be not 

 so slow, that the infinitesimal period over which our observations can 

 extend gives us no clue as to its direction. 



From the standpoint of morphology it seems to me that we are 

 free from all these difficulties. We take as our units not individuals 

 but species, and if the doctrine of descent be admitted, we must infer 

 that specific characters have themselves been of importance in the 

 struggle for existence, or have been inextricably bound to others 

 which have. If specific characters then are our material, certainly 

 they offer a sufficiently wide field to work on and give us a unit small 



