THE DETERMINATION OF DOMINANCE. 325 



tyrosin be present. There is in this production of melanin pig- 

 ment exactly the requirements of the Neo-Mendelian factorial 

 hypothesis, as suggested by Cuenot (1903) and developed by 

 many workers in the last few years — two things which must be 

 brought together to produce a definite end result, or if one be 

 absent then the one present is incapable of producing the custom- 

 ary result, i. e., pigment. 



In plants, it has been shown by Correns, by Bateson and his 

 students, and others, that two factors are necessary to produce 

 color; and in Antirrhinum, Wheldale (1907) and Wheldale, Mar- 

 ryat and Sollar (1909) have a considerable body of evidence to 

 show that there are specific chemical substances — chromogens — 

 to be oxidized to produce the colors of the flowers. Bateson 

 (1909) has clearly stated that this does not mean that any sub- 

 stance, as a specific substance, or entity, is segregated in the 

 germ cells; it simply means that a capacity for the production 

 of either one or the other, or both of them, must be present in 

 order to produce definite results. This capacity, as far as my 

 experience goes, seems in all cases to be a property of the whole 

 organism, which during ontogeny, manifests itself epigenetically, 

 now in one part or character, now in another, and a very simple 

 change at the start may conceivably result in extensive end 

 results. I am, while fully convinced of the truth of the behavior 

 of characters Mendelianwise, equally unconvinced, as the result 

 of ample experience, that the process is in any way due to a 

 particulate composition of the germ-plasm, or to any sort of 

 preformation, and as far as I am aware, no Neo-Mendelian has 

 thus far attempted to form any conception of what it is in the 

 germ cells that is productive of the observed results, and only 

 rarely do we find such concrete statements of preformation as 

 that made by Davenport. 



Castle (1909, page 68) expresses more accurately the current 

 Neo-Mendelian conception of these factors: "In what form, it 

 may be asked, are we to suppose that the various assumed factors 

 exist. Do they occur as so many different substances lying side 

 by side but unmixed in every reproductive cell? . . . 



"It is, we think, not necessary to suppose that there exist in 

 the minute germ-cell as many complex organic substances as 



