PiEACTIOXS AND PRODUCTS IN InTEESPECIFIC CkOSSES 173 



the extracted types coming from the crossing of signaticollis and diversa, the 

 degree of fertility is as high or higher than in either of the parent species. Com- 

 parison of the data shows clearly that there is relatively higher fertility in the Fg 

 extractives, showing that there has heen a change in the nature of the gametes 

 by the reaction of crossing, and inhibition of uniform fertilization and develop- 

 ment is removed by passage through the F^ reactions. A point of interest 

 arises in this connection, namely, the increased interfertility between these 

 extracted Fj stocks and the parent strains. 



The reactions in these crosses are of monohybrid or trihybrid types, and the 

 extracted products show no change in any detectable attributes or qualities, so 

 that the reason why the extracted products of crossing have higher interfer- 

 tility is a problem for further study. There clearly is some change in the 

 aametes as a result of the passage through the F^ reactions which in some way 

 removes the barriers to complete fertility in the matings of these otherwise dis- 

 tinct types. 



I was further interested to see what the data would show with respect to the 

 fertility of the Fg extractives coming from different crosses. I had compiled 

 from my mating-records data of the crossing of Fg homozygous-acting extrac- 

 tives, and although the crosses of this sort have not been numerous, the data 

 derived as the by-product of other experiments show clearly that the fertility 

 is strikingly lower than in the matings of Fo extractives from the same type of 

 species cross. 



Out of these data three interesting general points come : (1) the demonstra- 

 tion that alternation in the degree of fertility follows directly any change in 

 the value of the Ac determiner; (2) the demonstration that Fg extractives have 

 entire interfertility between themselves and with the parent species; (3) the Fj 

 extractives from different species crosses, but of the same extracted type, do not 

 have the same interfertility as extractives from the same type of cross. 



In the crossing of the species L. decemUneata, oblongata, and muUitcBniata 

 the Ac determiner values are nearly the same and culture of these species affects 

 no appreciable change in the Ac value; nevertheless, the fertility at the onset of 

 any stock direct from nature is decidedly lower than it is later in its history. 



As in the first three species, the fertility between these three never has reached 

 in experience the same ratio of progeny production from the total matings as it 

 does in the normal stocks under cultivation, but the F, and Fg extractives show 

 precisely the same relation that was found in the first group of species, showing 

 that the principles at work are the same. 



The fertility between these two groups of three species is very low, and thus 

 far the only series of crosses that has been at all successful are the two lines 

 from the crossing of decemUneata and diversa, described in the early portion of 

 this chapter, and which gave such unusual results. Only two have given adult 

 progeny out of many trials. No other successful crossing between the two 

 groups has been attained. This has been attempted, but there seems an aversion 

 of the two series to one another, so that copulation can not be obtained excepting 

 at the climax of reproductive activity, and then is best attained if the antennae 

 are cut off or covered with neutral gmm, so that the sex-reactions depending 

 upon the odors of the species secretions comes feebly or not at all into play. By 

 subterfuges of one sort or another I have been able to make all the possible com- 

 binations of these two groups of species, able to obtain matings and fertilized 



