438 Transactions of the Society. 



plumage in a score of ruffs, the number of vertebra; in a hundred 

 fishes of the same species, and so on. We register the observed 

 differences. 



It soon becomes plain, however, that analysis of our data i& 

 necessary, if we are to avoid fallacy. We must try to sift out 

 peculiarities which are associated with age and with sex, or are 

 directly due to peculiarities of nurture. It is obvious that immature 

 herrings must be compared with immature, and that we must not 

 mix up the ruffs and the reeves, drones and worker-bees. More 

 difficult, however, is it to separate off those peciQiarities which 

 can be experimentally shown to be individually acquired modifica- 

 tions, directly due to peculiarities in nurture (whether nutritional, 

 environmental, or functional). Many crabs are profoundly changed 

 by being parasitized by Sacculina and other Ehizocephala, and a 

 conclusion as to variability in crabs is vitiated by mixing up the 

 parasitized with the normal. An organism dwarfed by lack of 

 food or lack of space for exercise, such as the freshwater snails 

 studied by Semper and De Varigny, is in a different category from 

 a normal dwarf appearing in a family with no dwarfs in its recent 

 lineage. The much cut-up leaves of the freshwater buttercup in 

 the swiftly flowing water, one of the examples Lamarck gave of 

 the direct results of environmental influence, are not to be placed 

 beside the laciniate leaves of the variety of Chclidonium majns^ 

 which cropped up a long time ago in an apothecary's garden in 

 Heidelberg, and has been breeding true ever since. These directly 

 induced, exogenous modifications Darwin" called "definite varia- 

 tions " ; they are badly called " individually acquired characters " ; 

 they are best called " somatic modifications." As there is not 

 at present any convincing proof of the transmissibility of these 

 somatic modifications as such or in any representative degree, they 

 must be left, in the first instance, out of account in our inquiry 

 into the origin of the distinctively new. They may be of great 

 import for the individual, but if they are not transmitted they 

 cannot be of more than indirect importance to the race. It does 

 not follow, however, that a changeful environment may not be an 

 originative factor in evolution. 



When we subtract from the total of observed differences those 

 that can be shown to be modifications, when we also eliminate 

 the peculiarities associated with differences of age and sex, the 

 remainder are for the most part (in proportion to the success of 

 our subtraction) what are called variations — inborn not acquired, 

 intrinsic not extrinsic, blastogenic not somatogenic, endogenous 

 not exogenous, arising from the constitution of the germ-cell not 

 impressed from without, expressions not dints. Some of them at 

 least are very transmissible, and it may be said that these con- 

 stitute the raw materials of evolution. 



§ 3. The next step is to inquire whether all the inborn varia- 



