370 



SCIENCE. 



[N. S. Vol. XXII. No. 560. 



sometimes even, though rarely, revert to the 

 ancestral form. When varieties are crossed 

 with the ancestral species the characters of 

 the species are dominant in the hybrids, so 

 that the hybrid appears like the species. De 

 Vries finds, accordingly, that many cases of 

 apparent reversion, or atavism, in the offspring 

 of a variety of plant are due to the accidental 

 cross pollination with the ancestral species; 

 the ' reverted ' forms are really hybrids. This 

 phenomenon is called by de Vries ' vicinism ' 

 — ^the false atavism being due to the presence 

 of the species in the vicinity. In addition to 

 this false atavism there can be distinguished 

 a true one. For, although in retrograde varie- 

 ties the quality that has dropped out may 

 become wholly lost, it may, on the other hand, 

 be merely latent and may reappear, usually 

 suddenly, in some individual. At this point 

 de Vries introduces the idea of positive varie- 

 ties as opposed to negative ones. These are 

 characterized by the addition of a quality 

 which had been latent. (The addition of a 

 new quality is, it will be recalled, in de Vries's 

 scheme, the origin of an elementary species. 

 But will it not be often impossible to say 

 whether a new appearing quality is truly new 

 or old?) 



In cross-breeding the contrast between 

 varieties and species still holds. When a new 

 species, which is characterized by a new qual- 

 ity, is bred to the parent form, its germ cell 

 bears a unit-character of which the parent 

 species offers no representative with which it 

 may be mated in the conjugation of the sex 

 cells. This sort of cross is called a 'uni- 

 sexual ' cross (unbalanced or unsymmetrical 

 cross would have been less ambiguous). 

 Varieties, on the other hand, have the same 

 characters as the parent species, only one of 

 them is latent. This latency of the charac- 

 teristic does not prevent its union with the 

 corresponding patent characteristic of the 

 parent species. The offspring of a unisexual 

 cross are apt to have the various characters of 

 one or the other parent intact, fully developed, 

 side by side. These first generation hybrids 

 are all of the same type; in general aspect 

 they are intermediate between the two species 

 and this intermediacy persists when the hy- 



brids are inbred. In the so-called bisexual 

 crosses, on the other hand, the first hybrid 

 generation is said to resemble the parent spe- 

 cies. When such hybrids are inbred a segre- 

 gation of the various characters in different 

 individuals appears. This segregation occurs 

 in accordance with Mendel's law, and that law 

 is applicable only to bisexual (balanced) 

 crosses. 



The fourth section treats of ' ever-sporting ' 

 varieties and is the most novel and suggestive 

 in the book. It is essentially experimental, 

 yet one feels that the results gained are tenta- 

 tive. Ever-sporting varieties are defined as 

 ' forms that are regularly propagated by seed, 

 are of pure and not hybrid origin, but which 

 sport in nearly every generation.' Of such 

 varieties two types are recognized : ' poor ' 

 races or ' half ' races, and ' rich ' races or 

 ' middle ' races. In the former a sport trans- 

 mits its peculiarity to only a small percentage 

 — one per cent, to three per cent. — of its 

 progeny. In the latter a transmission to 

 twenty-five per cent, to fifty per cent, of the 

 offspring may occur. However, the two sorts 

 of races are not so sharply differentiated as 

 these figures would indicate, but naming the 

 extremes will accentuate the fact of varia- 

 bility in the transmissibility of sports, in 

 plants. The same thing is found in man. 

 Polydactylism is in some families strongly 

 transmitted, in others less strongly, in others 

 almost not at all. De Vries was unable to 

 establish a race of five-leaved crimson clover, 

 whereas he quickly got a race of five-leaved 

 red clover. Monstrosities behave like other 

 variations, showing both poor and rich races. 

 In the further development of the subject 

 de Vries is led to explain cases of functional 

 adaptation in plants on the ground that two 

 types are always present in species showing 

 these adaptations, and that ' during their ju- 

 venile stage a decision is taken in one direc- 

 tion or the other.' He even goes so far as to 

 ascribe the difference in height of some plants, 

 according as they occur in rich or poor soil 

 to a dimorphic tendency toward, one or the 

 other stature. This conclusion requires sta- 

 tistical proof before it will be generally ac- 

 cepted. 



