192 THE FUNCTION OF SEX 



this problem, we must bear in mind that sexual reproduction is 

 almost universal; whereas, to judge by facts that have been 

 actually observed, not merely imagined, Mendelian reproduction 

 is one of the rarest things in nature. Speaking practically, it 

 occurs only when mutations are reproduced ; and all the evidence 

 indicates that mutations occur relatively seldom, and, even then, 

 are very seldom preserved, except when the selection is artificial. 

 Mendelian reproduction, therefore, is, in effect, unknown in 

 normal that is, z>z/nz-varietal breeding. Natural varieties very 

 rarely cross, and only rarely display Mendelian traits when man 

 crosses them. Such traits are common only when artificial 

 varieties, in which mutations have been accumulated by man, are 

 crossed by man. They seem common to experimental workers 

 only because the materials for thought to which they have limited 

 themselves have been obtained almost exclusively in this latter 

 way. I think, therefore, that no one who has been accustomed 

 to found his judgments, not on the exception but the rule, will 

 hesitate to conclude that Mendelian reproduction is a variety of 

 sexual reproduction, and that the converse is as absolutely untrue 

 as if we had declared that mammals are a variety of dogs. 



317. If we found our judgments on verified evidence alone, 

 the conclusion seems irresistible that Mendelian ' inheritance ' is 

 a human creation, and that the right interpretation of the facts 

 appears to be, that nature treats mutations, when man interferes 

 and presents them to her, as sexual characters. She is not a 

 conscious selecting agency, and they resemble the sexual characters 

 and differ from the fluctuations which she normally meets, in that 

 they constitute wide differences between mating individuals. But 

 her treatment of them is more or less uncertain ; hence, as 

 compared to sexual' traits, their imperfect alternation and their 

 tendency to blend in greater or lesser degree. It appears, then, 

 that Mendelian characters are nothing other than non-sexual or 

 semi-sexual characters abnormally reproduced in the mode that 

 sexual characters are normally reproduced. It follows that experi- 

 mental workers have been engaged in investigating, not heredity 

 in general, not even the function of sex, but only certain anomalies 

 of sexual reproduction which occur under conditions of artificial 

 selection and crossing. They are the unhappy victims of a vast 

 practical joke unconsciously played by the human breeder. 



" The earth hath bubbles, as the water has, 

 And these are of them." 1 



^'Macbeth," i. 3. 



