I9I0-] THROUGH HYBRIDIZATION. 283 



never become fixed. However, in the work of Godron and Naudin, 

 their near discovery of Mendehan segregation was not due to a 

 dehberate consideration of the various characteristics as units, but 

 rather to the fact that several of the forms which they used in their 

 cuhures, differed from each other by single unit-characters, as 

 exemplified for instance by the purple color of stems in Datura 

 Tatula, contrasted with the green stem of D. Stramonium, or the 

 usual prickly fruit of the Daturas contrasted with the smooth fruit 

 of a var. " capsulis mermibus." 



The Mendelian method of following single characteristics 

 possessed by the parents, not only into their F^ progeny, but also 

 through the second, third, and later generations, brought to light a 

 regularity of behavior which has served to shift the stress from the 

 simple combination phenomena involved in hybridization, to the 

 phenomena of separation and recombination of such elementary 

 differences as existed between the two parents. 



The result of this important innovation of method has been 

 to demonstrate beyond a peradventure, that many of the distinguish- 

 ing characteristics of adult plants and animals are predetermined by 

 corresponding differences in the constitution of the germ-cells ; that 

 these differences may be of an elementary character, capable of 

 separation into different germ-cells; that when two parents used in 

 any cross, differ by such elementary characters, half the resultant 

 germ-cells have the capacity to produce any given elementary char- 

 acter of the one parent, the other half possess the capacity for the 

 production of the corresponding or alternative characteristic of the 

 other parent; that as a rule such unit-characters are wholly inde- 

 pendent from one another and capable of rearrangement in every 

 possible combination with one another ; and, finally, that it is purely 

 a matter of chance, which available type of sperm shall fertilize any 

 given Qgg. 



The separation of the unit-characteristics into different germ- 

 cells in every possible combination with other characteristics gives 

 the power in many cases to recognize all the unit-dift"erences which 

 served to distinguish the two parents. By the study of the hybrid 

 progenies we are thus given an insight into certain phases of the 



PROG. AMER. PHIL. SOC., XLIX. I96 S, PRINTED SEPTEMBER 6, I9I0. 



