Eeactions and Products in Interspecific Crosses 17? 



due to this cause rather than to the defective reproductive apparatus. Nor is 

 the fact that the gonads on examination have not produced gametes proof of their 

 sterility, because I have seen examples where there was not the least activity 

 until I had made some trivial change in the environmental complex. The repro- 

 ductive process is so delicately balanced in the organism, so tied with diverse fac- 

 tors within and without, that the problem of sterility presents endless complica- 

 tions to the experimenter. 



In this investigation fertility and sterility have been only incidentally studied 

 in the practical aspects of the breeding operations ; nevertheless, as a by-product 

 from these routine crossings of species, data are derived showing that the prob- 

 lems of this character are entirely matters of factorial causation and that the 

 production of conditions of fertility or intersterility are the direct result of the 

 reactions incident to the origination of the stable type, which, if it were found in 

 nature and iirst seen by the taxonomist, would have been a species. Isolation in 

 reproduction and intersterility have long been and still are by many considered 

 the distinguishing mark of species, but I can not see any a priori reason why this 

 is necessarily so. Specific distinctness in the non-living is solely the possession, 

 by the contrasted materials, of constant and characteristic attributes and quali- 

 ties, manifested to our perceptions, such that they may by us be separated one 

 from the other. They may be specifically distinct, yet interact with violence, ease, 

 or not at all. The capacity for interaction is not a criterion of specific distinct- 

 ness. In organisms fertility or sterility in its many manifestations is only the 

 capacity of the specific masses to react, and the mere fact that they act readily 

 and uniformly when combined, with difficulty, or not at all, has no bearing upon 

 their specific distinctions, which in each are proper thereto and constant at all 

 times, irrespective of the capacity for interactions. Enough has been presented 

 at this point to show the trend of ideas with reference to these problems and the 

 substantial support that comes to them from these experiments in the crossing of 

 species. Interspecific sterility in animals has been regarded as an indelible char- 

 acteristic of specific distinctness; but as Darwin was forced to admit, and all 

 since have seen, this was not universally true. Darwin, however, thought that it 

 was true in nature, but could be removed by domestication. Cultural conditions 

 can alter the factors concerned in the production of sterility, within the limits of 

 fluctuation proper to the species, not beyond, and further change must come as 

 the result of the product of the reactions of interacting gametic systems, wherein 

 the factorial nature of the system is itself changed. I have reared diversa and 

 decemlineata for many generations and seasons in the same conditions, but the 

 degree of fertility is no higher now than at the beginning. Others have changed 

 within the limits of fluctuation of some specific and efficient factor, not beyond, 

 and perfect fertility is not in experience produced in any instance by the condi- 

 tions of culture. 



THE PROBLEMS OF THE NATURE OF SPECIES CROSSES. 



De Vries has clearly crystallized the general opinion so long prevalent that the 

 crosses of species are diff'erent from the crosses of varieties in principle, in his 

 now familiar statement that there are two main types of crosses — the l)isexual or 

 Mendelian and the unisexual — in which the chief criterion is that the bisexual 

 gives in Fg segregation into different groups with interchanges of the character- 

 istics of the parents, some of which breed true as varieties, while the unisexual do 



