IN THE TROPICS. 
and distinguished them as a group from the “ true hybrids ” 
which follow Mendel’s Law. The group in question is 
characterized as followsThe first generation of “false 
hybrids ” may be uniform or di- or poly-morphic. As a rule 
they breed true ; but they may show “ splitting.” Such 
“ splitting ” when it occurs is not into equal but into unequal 
parts. 
This classification is distinctly unsatisfactory, and it is 
not repeated in de Vries’ later papers. 
De Vries has also studied the behaviour of a number of very 
variable characters on crossing, and concludes that these 
also obey Mendel’s Law in fundamentals. In the case of 
crosses between tricotyledonous “half—” and “mid-races” 
he shows with a high degree of probability that Mendel’s 
Law is followed in this case too, in spite of the existence of 
transgressive variability. 
De Vries believes that by the aid of Mendel’s Law he has 
solved the problem of the difference between species and 
varieties, or rather, of that between specific characters and 
varietal characters. According to the mutation theory, 
varietal characters arise by the “retrogressive” or by the 
“degressive” method of mutation, and follow Mendel’s Law;* 
specific characters which arise by progressive mutations do 
not. Thus, in terms of his theory of intracellular pangene¬ 
sis, de Vries believes that the pairs of characters which on 
crossing obey Mendel’s Law are always represented by the 
same pangen in different states, active and latent respec¬ 
tively in the most usual case. Characters which havearisen 
progressively do not follow Mendel’s Law, but obey the law 
of unisexual crossing ; that is to say, they produce blended 
progeny which breed true. This being the case, the long- 
established view that varieties are perfectly fertile on 
crossing, whereas species when crossed show a diminution 
* In such a case the phylogenetically older character is represented by 
the active pangen and is dominant. Bateson and Saunders, however (#). 
