94 MR. W. BATESOX ON COLOUR-HEREDITY [May 25, 



As we have also seen, the colours taken collectively follow simple 

 expectation ; F^ x F^ giving approximately 3 colovu-ed to 1 albino, 

 and F^X albino giving approximately equal numbers of each. 



As to the frequencies and valencies of the particular colours 

 nothing can be said with much confidence as yet, beyond the 

 statement that F^ gives off albino gametes about equal in number 

 to the various coloured gametes collectively. In a discussion of 

 this subject, Professor Weldon (25) has suggested that an average of 

 one albino in nine might have been expected. I can see no 

 reason why this proportion should be impossible in nature, from 

 Fj X F,. Its occurrence would, however, be remarkable and raise 

 some important problems in gameto-genesis. So far, however, it 

 has not been recorded. Professor "Weldon is in error in stating 

 (25. p. 34) that I have already dealt (4. p. 52) with such a case 

 of 1 albino in 9. The case in question was that of Antir7^him(,in, 

 where de Vries obtained from F^ x F^ four forms in the proportion 

 9:3:3:1, the one being the white, which therefore occurred 

 in the proj)ortion of 1 in 16. This is the proportion Mendel 

 himself conjectured might be found in a case of resolution, but I 

 do not gather that he had actually observed such a case. 



!No case of resolution has yet been sulficiently studied for us to 

 speak with any confidence as to the ratios of the gametes or the 

 nature of the process of resolution. Tschermak has had cases of 

 1 recessive in 4, after resolution. In poultry I have had cases 

 somewhat similar, to be described hereafter. 



In apparently all recorded cases of resolution some gametes of F^ 

 carry the compound character unresolved. It is not at all easy to 

 suggest a scheme which shall fit both the observed facts of resolu- 

 tion and those of cell-division. For example, suppose the gametes 

 of F^ to be 50 per cent, albino, 50 per cent, variously coloured, if 

 segregation were complete. Let us consider the coloured gametes 

 separately, and for simplicity assume there are only three kinds of 

 them, viz. the unresolved grey, black, and yellow, the two lattei' 

 being hypallelomorphs of grey. It is then clear that in whatever 

 numbers the three types are each represented, so long as their 

 sum equals the total of albino gametes, there must be more 

 black character in any black gamete, and more yellow in the 

 yellow gamete, than in any grey gamete ; or there must some- 

 where be a cell-division in which a part of the yellow and a part 

 of the black have been lost. If, for instance, the hybrid bore 

 gametes in the proportions 



2 grey (zi: black + yellow), 1 black, 1 yellow, 4 albino, 



we recognise that unless the blacks and yellows carry double 

 portions of their respective colours, part of the colour originally 

 introduced into Fj has been lost. Such doubling is not altogether 

 inconceivable, though until histological methods are made applic- 

 able to these questions of gameto-genesis the possibibity can hardly 

 be tested. We note as a fact favourable to such a view, that the 

 visible amount of pigment in a black or a yellow zygote is far 



