Presidential Address. 9 



In such a classification of characters, there is nothing to correspond with 

 de Vries' division of variations into mutations — those fortuitous variations 

 which alone, according to de Vries, play a part in Evolution — and fluctuations, 

 or ordinary variations, which do not play any part in Evolution, for it does 

 not appear to me that the existence of any such fundamental distinction has 

 been demonstrated, or even rendered probable. 



De Vries adduces evidence to show that artificial selection is unable 

 to accumulate certain variations (fluctuations) beyond a certain limit, and 

 argues that Natural Selection would be equally impotent. He also points out 

 that when the artificial selection is stopped, the organisms soon show 

 regression to their previous state — no permanent change having been 

 brought about by the artificial selection. 



To this it must be remarked, 1 that in reasoning from the data afforded 

 by artificial selection, it has to be borne in mind that artificial selection 

 selects definite isolated characters — the result being a thoroughly abnormal 

 race of individuals — in which, by the exaggeration of the selected features, 

 the normal physiological balance of parts in the organism has been disturbed. 

 When Nature selects, on the other hand, she has the whole organisation of the 

 creature firmly in her grip the whole time. The individuals are of necessity 

 physiological, not pathological, 2 and there is no obvious reason why under 

 those conditions there should be any limit to her power of producing change. 



When artificial selection ceases there is a regression towards the original 

 condition — -in other words, a return towards the physiological from the 

 pathological, which is surely not surprising when one considers the immense 

 weight of hereditary tendency on its side. Unlike artificial selection, Natural 

 Selection does not cease ; organisms are kept in Nature's merciless clutch the 

 whole time, and there is therefore no regression. 



The Conception of Units in Biology. 



I confess that my suspicions are at once aroused by all investigations 

 which deal with isolated "individual" characters apart from the general 

 organisation of the animal. There is an insatiable desire in the human mind 

 to resolve a complex into the simple units of which it is built up, and it is of 

 course to the carrying out of this that we owe the triumphs of the analytical 

 method in the most varied departments of Science. In regard to living 

 substance, it seems a very natural idea that it is built up of ultimate units. 



1 Poulton expresses the same criticism in his Darwin and ike Origin, p. 279. 



2 No weight can be attached to the experimental failure to increase the quantity 

 of sugar in the cells of the sugar beet beyond a certain limit. Even sugar is a poison if in 

 sufficient quantity. 



