122 RETROGRESSION 



the sum of adaptation, leaving only the other half to the direct 

 action of Natural Selection. Yet all these great effects result 

 solely because in every structure of every species the tendency to 

 vary retrogressively is so much stronger than the tendency to vary 

 progressively that retrogression is checked only by selection. 

 Obviously this tendency of retrogressive variations to preponderate 

 over progressive variations is highly adaptive as adaptive as 

 spontaneous variability itself. The question then arises whether 

 it is an actual adaptation or merely something which has arisen 

 accidentally during the course of evolution. We saw that the 

 regulated variability which distinguishes all forms of life war 

 beyond reasonable doubt, maintained, and controlled by Natural 

 Selection. The tendency to retrogression is only one of the ways 

 in which variability is regulated. We have, therefore, every reason 

 to suppose it is no chance accompaniment of life, but an adaptation 

 ranking in universality and importance with spontaneous varia- 

 bility, insusceptibility, and recapitulation. In other words, we 

 have reason to believe that it is a LA vv that has resulted from the 

 selection of germ-plasms which, while tending to vary all round the 

 specific mean, yet tended, on the whole, to vary retrogressively. 



202. The problem of retrogression cannot be approached by 

 any of the laboratory methods and therefore tends to be ranked 

 by some minds among the questions of * philosophy.' Moreover, 

 the very simplicity of the device by which nature achieves such 

 great results is apt to awaken incredulity. I can only ask the 

 reader, while bearing in mind the facts of adaptation and also the 

 fact that both useless progressive variations and characters that 

 have become useless tend to disappear in the absence of direct 

 selection, to try here again to conceive interpretations other than 

 those I have suggested. Judging from my own experience, I 

 think he will fail. 



203. We conclude, then, that the great mass of retrogression 

 we observe in nature is not due directly to Natural Selection that 

 is to reversed selection. Since, as a rule, environments change 

 gradually, it occurs under conditions which afford no scope to 

 reversed selection. It is not to be supposed, for example, that the 

 apteryx suddenly abandoned habits of flight. While the struc- 

 tures and instincts which especially'fitted it for a terrestrial life were 

 undergoing progressive evolution, its wings were still helpful in 

 sustaining existence. But they became continuously less useful, 

 were less and less maintained by selection, and so at last 

 underwent retrogression. Sometimes, however, we may observe 



