586 Species According to the Theory of Mutation. 



course, any given monohybridic cross can only l)elong 

 to one of these two groups. If it is bi-sexual and be- 

 haves in a Mendehan fashion we may immediately infer 

 that tlie two parents are to be considered as varieties.^ 

 If it is uni-sexual they are elementary species, of which 

 the one must have been derived from the other. 



The di-polyhybrids are mongrels whose parents differ 

 from one another in respect of two or more elementary 

 characters. Two cases must be distinguished. Let us 

 confine ourselves first to the dihybrids. In some cases 

 the two points of difference may belong to the same 

 category and therefore follow the same laws in the crosses 

 and their products. If each of them, considered by itself, 

 would lead to the conclusion that the parents were related 

 as varieties, the same conclusion will obviously hold good 

 for the combination. So, for instance, is Pa paver sonini- 

 fernm polycephahun Danehrog to be regarded as a vari- 

 ety ; so also, if it is allowed to judge by analogy, Calliopsis 

 tinctoria piimila purpurea (Vol. I, p. 197). For the same 

 reason the compound colors, which may be split into their 

 components by means of crossing and can be recon- 

 structed out of these, fall within the category of varietal 

 characters. 



Similarly forms, of wdiich one has arisen from the 

 other by two successive mutations in the progressive di- 

 rection, and whose crosses, therefore, conform to the laws 

 of uni-sexual unions, would most certainly have to be re- 

 garded as elementary species. 



^ And this independently of the nomenclature chosen. For in- 

 stance according to the principles enunciated above, Chelidonium 

 laciniatiim Mill, will have to be regarded as a variety of C. majus, 

 even when this more convenient name is retained. See Part I, p. 65. 

 And from a practical point of view it would be very desirable to drop 

 the attempt to correlate the nomenclature with the ever-changing 

 systematic conceptions. 



