30I R. Ruggles Gates. 



Species which bear the same qualities in both the male and 

 female sex-cells, and whose reciprocal hybrids are therefore alike, 

 are designated isogamous ; while species whose functional male and 

 female cells are unlike in their latent capacities are called hetero- 

 gaiiious. Thus O. Hookeri, O. Cockerelli and 0. strigosa are found 

 to be isogamous as is also 0. Lamarckiana ; while O. biennis, O. 

 biennis Chicago, O. cruciata and O. 7Huricata are heterogamous. 



In all these cases there is nothing resembling the Mendelian 

 recombination of many independent characters, but the various 

 hybrids remain constant and uniform in later generations, except 

 in an occasional character such as flower-size. The hybrid 

 types agree in their main essentials for various crosses, and are 

 therefore given names : — rubiennis, conica, gracilis, rigida, etc. 

 Thus 0. inuricata x O. Hookeri, O. muricata x O. Cockerelli, 0. 

 Hookeri x O. Cockerelli and O. Cockerelli x 0. Hookeri, etc., all 

 give the rigida type, differing from each other only in minor 

 features. By such parallel series of crosses the character of the 

 "Pollenbild" and " Eizellenbild" of each species is determined, 

 and De Vries calls the process gamolysis. 



Among other cases in which a similar behaviour occurs may be 

 mentioned O. Hookeri x O. biennis, which gives an P, of the 

 rubiennis type, the latter splitting in later generations into rubiennis 

 and " Hookeri." O. biennis Chicago x O. Hookeri and O. cruciata 

 X O. Hookeri give the same result. This is explained by the 

 isogamous condition of O. Hookeri and the heterogamy of the other 

 three species. In O, Hookeri x O. biennis, e.g., the rubiennis hybrid 

 bears in its egg cells only the characters of Hookeri while the 

 pollen bears the segregated characters of both parents. Hence the 

 type of splitting observed. 



Further, not only these crosses but also the twin hybrids (Iceta 

 and velutina) and the equally extensive series of mutation crosses 

 (with the mutants), involving yet a third type of hereditary behaviour, 

 are all finally explained and harmonized in connection with the 

 theory of mutation. 



De Vries explains not only all this hereditary behaviour but 

 also the mutation phenomena in terms of his hypothesis of intra- 

 cellular pangenesis. Each pangen represents a special character, 

 and a pangen may be in (i.) the active, (ii.) the inactive, or (iii.) the 

 labile condition, pangens in the labile condition giving rise to 

 mutations. A mutation also consists in the passage of a pangen 

 from one condition to another, or sometimes in the addition of a 

 new pangen. Thus in O. mut. nanella the pangen for stature has 

 passed from the active to the inactive condition, in 0. mut. rubrinervis 

 this pangen is active, while in 0. Lamarckiana it is in a labile 

 condition. Hence Lamarckiana x nanella gives some dwarfs in Fj 

 while rubrinervis x nanella yields only tails in F^ but a varying 

 proportion of dwarfs in F^. 



I have only touched upon a few of the items in this remarkable 

 book, which boldly attempts to explain all the intricate breeding 

 behaviour in CEnothera. Though one cannot agree with all its 

 statements, yet the lucid explanations given make the work of great 

 value to all students of heredity and evolution. 



One other point to which reference may be made is found in 

 the recent work of Gates and Miss Nest^ Thomas, They have not 



