cussing the phenomenon of latency in the light 

 of the accumulated data. 



The appearance of hereditary characters 

 which are not traceable to the immediate an- 

 cestry offer the most difficult problems with 

 which the student of heredity must deal. 

 These characters may be recognizable as hav- 

 ing belonged to the more remote ancestry 

 of the form in question, i. e. } they are atavistic, 

 or they may be wholly new. Sometimes they 

 occur under known conditions, at other times 

 there is no clue to the causes upon which they 

 may depend. 



These bean hybrids which possess characters 

 not seen in either parent furnish good ex- 

 amples of latency and if we can determine 

 whence the new characters came we shall be 

 far advanced toward a correct conception of 

 latency. Taking, for instance, the hybrids 

 between the yellow and white and between the 

 brown and white, two characters may be recog- 

 nized as new, viz., the dark pigment and the 

 mottled color-pattern. The fact that the very 

 same characters appear in the hybrid off- 

 spring, no matter what the character of the 

 pigmented parent, leaves little doubt that both 

 of these new characters are traceable to the 

 white parent, and we should seem to be war- 

 ranted in saying that this white bean con- 

 tains latent purple and latent mottling. 

 Tschermak 10 takes just the opposite view, how- 

 ever, and would say that the purple mottling 

 is latent in the pigmented bean, and that the 

 white bean acts simply as a releasing agent 

 which allows the latent character to become 

 manifest. Bateson 11 also subscribes to the 

 same view. On Tschermak's hypothesis it 

 would appear to me a remarkable coincidence 



10 Tschermak, E., loc. cit. 



"Bateson, W., Saunders, Miss E. R., Punnett, 

 E. C., ' Experimental Studies in the Physiology of 

 Heredity,' Second Report to the Evolution Com- 

 mittee of the Royal Society, London, 1905. 



