VARIATION 417 



' tendency ' of the protoplasm ; but how are we to picture 

 the origin and continuance of meristic variations ? 



A separate consideration may be given to fertilisation as 

 a source of variation, a view prominent at one stage in the 

 development of Professor Weismann's theories. For a time 

 he was inclined to attach great importance to the mingling (or 

 amphimixis) of two sets of hereditary qualities as a pos- 

 sible source of novelties, but he afterwards attached more 

 importance to the influence that fluctuations in nutrition 

 within the body might have in inducing changes in the 

 germ-plasm or in inducing struggle among the analogous 

 hereditary items. In recent years the Belgian botanist Lotsy 

 has been a thoroughgoing champion of the variational signifi- 

 cance of fertilisation and has gone the length of maintaining 

 that all variation is due to crossing. There is ample experi- 

 mental evidence that novelties may be induced by crossing, 

 and this is not surprising when we remember that two very 

 complex systems, usually of diverse origin, become in fertili- 

 sation a unity that goes on in most cases to develop into a 

 harmonious life. On the other hand, Lotsy's attempt to 

 refer all variations to crossing is extreme. This is shown, for 

 instance, by the occasional occurrence of variations in 

 parthenogenetic lineages in which no father intervenes for 

 prolonged periods. Moreover, crossing can be of no avail 

 unless the two sex-cells that combine are different. If they 

 are different it must be by hypothesis because of previ- 

 ous crosses. Thus we simply push the problem back 

 and back to original differences which are left unaccounted 

 for. 



The problem before which we are bafHed is the origin of 

 the distinctively new, where the novelty is qualitative not 

 quantitative. Some would refuse to admit this distinction, 



