672 REPORT— 1901. 



Ireland the immediate result may be complete sterility. The tendency of some exotic 

 plants to ' sport ' after tliey become acclimatised is doubtless due to tbe fact that 

 their new habitat is uniisually favourable, their general vigour — so essential for 

 new developments — is increased, and, probably because certain groups of germ 

 units are constantly stimulated by the new food available, they give rise abruptly 

 or gradually to new and it may be unexpected characters. No one doubts that 

 the bodily vigour is liable to be impaired by fevers and other diseases, by changes 

 in the habitat, unsuitable food, rapid and unseasonable changes of temperature, 

 and the like ; hence it will not be surprising if further investigations prove that 

 changes in the soma, beneficial as well as injurious, are reflected in the germ-cells, 

 and thus indirectly induce variation. Moreover there are excellent reasons for 

 believing that the germ-cells are influenced by seasonable changes, such as moult- 

 ing in birds and changing the coat in mammals. In the case of pigeons, e.g., the 

 young bred in early summer are, other things being equal, larger and more 

 vigorous, and mature more rapidly, than birds hatched in the late summer or 

 autumn. But however sensitive the germ-cells may be to the changes of their 

 immediate environment, i.e., the soma or body in which they are lodged, there is 

 no evidence whatever that (as Buffon asserted and Darwin thought possible) 

 definite changes of the soma, due to the direct action of the environment, can be 

 imprinted on the germ-cells. By the direct action of the environment — food, 

 temperature, moisture, &c. — the body in whole or in part may be dwarfed, 

 increased, or otherwise modified ; but such changes only influence the germ-cells in 

 so far as they lead to modifications in their vigour and nutrition. They may 

 expedite or delay maturity, alter the length of the reproductive period, interfere 

 with the nutrition of the germ-cells, or retard the development of the embryo, but 

 they seem incapable of giving rise to definite structural or functional variations iu 

 the offspring. 



Intercrossing and Interbreeding as Causes of Variation. 



The belief was once common amongst naturalists that variability was wholly 

 due to crossing, and at the present day naturalists and breeders alike agree that 

 intercrossing is a potent cause of variability, and are unanimous in regarding 

 interbreeding as an equally potent means of checking variability. The opinion is 

 also general that intercrossing has a swamping influence ; that having brought 

 forth new forms it forthwith proceeds to destroy them. Darwin, when discussing 

 reversion, points out that intercrossing often speedily leads to almost complete 

 reversion to a long-lost ancestor, i.e., to the loss of recently acquired and the 

 reappearance of long-lost characters.' When, however, he comes to deal with 

 variability, he states that ' crossing, like any other change in the conditions of 

 life, seems to be an element, probably a potent one, in causing variability,' - the 

 offspring of the first generation being generally uniform, but those subsequentl}' 

 produced displaying an almost infinite diversity of character. As to the influence 

 of inbreeding, he says ' close interbreeding, if not carried to an injurious extreme, 

 far from causing variability, teuds to fix the character of each breed.' ^ 



These statements may be quoted in support of the very common belief that 

 intercrossing is both a potent cause of variation and of reversion ; that it produces 

 new varieties one moment and swamps them the next. Whether intercrossing 

 may be regarded as the immediate cause of variation or of reversion (it can hardly 

 be both) depends on what is implied by variation. Obviously, variation may be 

 either progressive or retrogressive, i.e., the offspring may differ from their parents 

 in having quite new characters or iu presenting ancestral characters, or in being 

 cliaracteriscd by traits neither new nor old, due to new combinations of characters 

 already recognised as belonging to the variety or species. When intercrossing 

 results in the restoration of old characters, we have reversion or retrogressive 

 variation ; when to new combinations of already existing characters like new com- 

 binations in a kaleidoscope, we have new variations of a non-progressive kind, 



' Animals and Plants, vol. i. p. 22. 

 - Ibid., vol. ii. p. 254. ^ Jhid., vol. ii. p. 251. 



