Mat 24, 1907] 



SCIENCE 



831 



in the table above, representing the F, of the 

 cross between ' Ne plus ultra ' and ' White 

 flageolet ' it is seen that one of the white 

 derivatives from this cross has the same 

 gametic composition as the ' Marrowfat.' 



Emerson attempted no explanation of the 

 different behavior of these beans, simply pre- 

 senting them as exceptions to Mendel's laws 

 or as evidences of the limitation of the use- 

 fulness of those laws in predicting the results 

 of hybridization. Practically all of the ex- 

 ceptional results obtained by him cease to be 

 exceptional when we cease to look upon the 

 products of his crosses as monohybrids with 

 respect to seed color. His second generation 

 hybrids were classifiable into four categories 

 instead of the three he expected, but his ex- 

 pectation was based upon the assumption that 

 the black, brown, white, etc., are unit char- 

 acters, and that the mottled hybrids were 

 simply mosaics or blends between the white 

 and the self-colored parents. The simple as- 

 sumption, demonstrated in my hybrids, that 

 the pigmentation and the mottling are distinct 

 unit characters, harmonizes his results per- 

 fectly, though the numbers with which he 

 dealt were too small for the satisfactory de- 

 termination of the agreement or disagreement 

 with the theoretical ratios of a dihybrid. 



In discussing the appearance of purple 

 spotting as a novelty in peas, Lock^' follows 

 Tschermak in referring the latent character to 

 the pigmented parent, saying that " On cross- 

 ing A(B) {B being latent) with ah we get: 

 F^ ABab (B latent having become B active)." 

 If instead he' had considered that Ab is 

 crossed with aB, the A producing the pigment 

 and the B aggregating it into spots, he would 

 get the same F^, namely ABab, but would 

 have avoided the difficulty of a capricious unit 

 which may be active or inactive under condi- 

 tions that can not be determined. He would 

 then have had no latent character in any 

 other sense than that it was invisible owing 

 to the absence of pigment. 



The explanation here ofiered is essentially 



" Lock, R. H., ' Studies in Plant Breeding in the 

 Tropics,' Ann. Roy. Bot. Oard. Perideniya, 2: 

 299-356. See p. 341. 



that presented by Cuenot for mouse hybrids, 

 in which one unit is assumed to give pig- 

 mentation and another to determine the color 

 which this pigment will exhibit. Cuenot con- 

 siders the various colors to be latent in the 

 albino and he is supported in this respect by 

 my hybrids, but I prefer not to call this char- 

 acter a latent •pigment but an active pigment- 

 changer. 



This reference of various colors to the ac- 

 tion of a pigment-changer requires that the 

 pigments upon which the various colors de- 

 pend shall bear some simple relation to each 

 other. I have made some preliminary studies 

 on the pigments of these beans and have 

 partially demonstrated this simple relation by 

 converting the yellow and brown pigments to 

 black by the use of alkalies but I have not yet 

 been able to reverse the process. It is easily 

 demonstrable that the black (dark purple) 

 bean contains anthocyan, and this gives a 

 simple explanation of the correlation between 

 black seed-coats and red flowers, observed by 

 Mendel and all other students who have 

 chanced to use black-seeded peas or beans. 



That the yellow, red, and black pigments of 

 animals are closely related is also well known, 

 and there can be no doubt that the 'latent 

 black' which Castle" reported in certain 

 albino guinea-pigs is to be interpreted exactly 

 as Cuenot's mice, the black being due to the 

 presence of a melanizer which is a unit char- 

 acter wholly independent of the pigment-pro- 

 ducing unit. The fact that half the gametes 

 of this individual carried the so-called ' latent 

 black' simply showed the animal to be heter- 

 ozygous with respect to this allelomorph, and 

 the extracted recessives which did not in sub- 

 sequent generations produce any black off- 

 spring could not do so for the simple reason 

 that the pigment-changing unit had acted in a 

 perfectly normal way and had been absolutely 

 separated out into the black offspring while 

 its recessive counterpart was segregated with 

 equal purity into the non-black. 



A very important consideration in this con- 

 nection is the frequency with which the new 

 character is atavistic. This shows the process 



" Castle, W. E., loc. cit. 



