170 The Mechanism of Evolution in Leptinotaksa 



The first group of causes of sterility is at present not open to profitable dis- 

 cussion, owing to the fact that in no case is it known what the agencies were 

 that produced the mechanical inability to copulate or inhibition of the gametes 

 to unite or even come into proximity. These are matters of existing structures, 

 wdth specific modes of function, of whose method of production entire ignorance 

 prevails. Apply interpretation as one will, nothing but a miserable array of 

 plausible arrangements of possible causes and effects is the result. It is not 

 even probable that these diverse inhibiting mechanical devices arose because of 

 their utility in keeping the species pure or aided in its origin, and so we must 

 pass them by until such time as there is real basis of discussion. 



Of the two remaining categories of agencies productive of sterility and 

 associated with the union and nature of the gametes themselves, direct experi- 

 mental investigation is possible. 



It has been recorded in many plants and animals that the fertilized zygotes 

 resulting from a cross begin development, progress normally up to a certain 

 point, then cease normal development, go on in abnormal development for a 

 time, and then die at approximately the same point. It is needless to cite 

 instances ; these are too well known to the readers of hybridology and experi- 

 mental embryology. 



In the crossing of these natural species I have had many indications that the 

 problem of sterility is entirely one, as far as the gametes are concerned, of their 

 factorial composition and of the position and action of these agents in the 

 gametic complex. A priori this might be asserted, but can experimental con- 

 firmation of it be obtained, investigated, and the role of specific agents deter- 

 mined as to their relation to the production of this phenomenon ? 



Uniformly in all of the species used the stocks fresh from nature do not cross 

 with the same ease or certainty as they do after they have been in the laboratory 

 under the same conditions for several generations. This has been noted in one 

 way or another for nearly a century, and Darwin makes much of it in his writ- 

 ings, but I do not know that anyone has made a serious investigation thereof. 

 With the species used I have had no difficulty in their breeding under the cul- 

 tural conditions provided for them, so that my problems are not those so often 

 mentioned in the breeding of wild organisms — of their refusal to breed under 

 domestication. I have no"t found any that would not reproduce when given the 

 proper food and conditions in the medium. Nevertheless, while I have had con- 

 stant success in the breeding of wild species in my laboratory conditions, the 

 crossing of these shows in all most decidedly that at the start the cross is not only 

 far more difficult to make, copulation is not so easily effected, and after copula- 

 tion, if fertilization is effected, the mortality in the developmental stages is 

 immensely greater than in the same type of cross in the same strain two or 

 three years later. Data upon this point is given in condensed form in table 23 

 which shows in the three species L. signaticolUs, diversa, and undecimlineata 

 the change in respect to the fertility of the matings in crossing of these three 

 during the first six generations of captivity. In the pure species cultures there 

 has been no appreciable change in the fertility of the species from the introduc- 

 tion to the end of the sixth generation, and the matings not producing adult 

 progeny are the same in the first as in the last generation ; nevertheless, in the 

 crosses there is in the matings made in the stocks fresh from nature much lower 



