1899] EVOLUTION AND CHANCE 359 



another example : it is a scale, but a scale that has been much divided 

 and very highly elaborated. Is it possible to believe that by a process 

 of mere addition a reptilian scale could become a feather ? The plants 

 from which our warden flowers are descended had themselves no 

 flowers. The petals are no doubt leaves that have gathered in a 

 whorl and changed their colour. But here again we have a change 

 which cannot possibly be represented in terms of more or less. 

 Parenthetically I may remark that I leave out of consideration the 

 Lamarclrian view. The idea that the crawling of bees or other insects 

 over plants, or anything in the environment can have produced flowers 

 is too great a strain on the credulity of an ordinary man. To me 

 every acquired character is, as Weismann has put it, " simply the 

 reaction of the organism upon a certain stimulus." Two more instances 

 and I will leave this part of the subject. The whalebone that hangs 

 from the whale's palate is a fresh development. It cannot be accounted 

 for on Prof. Welclon's principle. In some fishes the air-bladder has been 

 turned into a lung and an entirely new connection has been formed 

 between it and a pair of aortic arches. Here, as in the preceding case, 

 we have a new departure. 



I am quite aware that if we hold that no discontinuous variation 

 ever leaves its mark, but is swamped by inter-crossing, there is no 

 such thing as a new departure. But discontinuity is a matter of 

 degree. If we say that only infinitesimal variations can survive, we 

 stultify ourselves, for such variations must be too small for natural 

 selection to work upon. We must, therefore, recognise some amount 

 of discontinuity, and the only question is, how much ? I cannot here 

 discuss the point at any length. But if, for instance, the growth of 

 whale-bone was at the outset very minute, it is difficult to see how 

 it could have been of any use, or have had selective value. In colour 

 we know the shifts are sudden. There is a sudden appearance of 

 albinism. Young domestic pigeons are sometimes quite different 

 in colour from either parent. Wild-flowers when planted in a garden 

 will often show some sudden conspicuous variation. However small 

 the bloodvessels connecting the aorta with the swim-bladder turned 

 into lungs, the connection at any rate must have been complete at the 

 outset. For what would be the use of a vessel that communicated 

 with nothing ? However we may seek to minimise it, we must 

 recognise some discontinuity, and in some cases some considerable 

 amount of it. That being so, we recognise new departures. And if 

 we recognise new departures, we cannot hold that evolution depends 

 only on deviations from a mean. 



If the law of chance fails us in many cases, where are we to 

 look for guidance ? I believe that Eimer has discovered where it is 

 to be found. In his "Organic Evolution," p. 52, he says: " I assume 

 with him (Nageli) that the conditions for a progress towards the more 

 complex, and towards division of labour, exists in the fact that 



