596 Spcxics According to the Theory of Mutation. 



These causes may fall into two entirely different cat- 

 egories ; on the one hand, they may consist in insufficient 

 systematic knowledge, on the other, in the insufficient num- 

 ber of experimental crosses. With regard to the former 

 point it should be remembered that, although the sys- 

 tematist frequently takes latent characters into considera- 

 tion, it is obviously by no means always possible to de- 

 cide on systematic grounds whether a character which 

 we do not see is really absent or exists only in a latent 

 condition. Nevertheless, latency is often regarded as a 

 retrogressive metamorphosis and therefore as the mark 

 of a variety: whereas absence is considered as a phylo- 

 genetically older step and therefore as a specific character 

 (see above p. 71). 



From this discussion we see that we may cross a plant 

 in which any given character is active either with one in 

 which the internal factor for this character is absent, or 

 with a species or variety in v/hich it is present but in a 

 latent or inactive state. Externally there is no difference 

 between two such crosses, but fundamentallv thev are 

 exactly opposite, and therefore it is to be expected that 

 their results will differ. The cross, active X absent is a 

 uni-sexual union and will presumably lead to a halving 

 of the external characters of the parents in the hybrid, 

 whereas the cross active X latent is a bi-sexual one, and 

 follows Mendel^'s laws, at least in ordinary cases. ]\Iany 

 paradoxes, which at present seem to negative the parallel 

 between systematic and sexual affinity may perhaps be 

 explained by more exact investigation on these lines. 

 FocKE gives the following cases as instances:^ "Silene 

 vulgaris, and S. luaritinia, Capsella ruhella and C. bursa 

 pastoris, Phaseolus vulgaris and Ph. inultiflorus, or the 



* FocKE, loc. cif., p. 448. 



