57(j THE PROBLEMS OF HEREDITY AND THEIR SOLUTION. 



zygote ABiovms its j^ametes, these arc not individually bearers nierel}'^ 

 of A and ^, but of a number of characters tliemselves ag-ain integral, 

 which in, say ^1, behaved as one character so long as its gametes united 

 in fertilization with others like themselves, but on cross fertilization 

 are resolved and redistributed among the gametes produced by the 

 cross-bred zygote. 



In such a case we call the character ^1 a compound allelomorph, and 

 we can speak of the integral characters which constitute it as hypallelo- 

 morphs. We ought to write the heterozygote {AuVA"...) B and the 

 gametes produced bv it may be of the form ^1, A! ^ A!\ ^1'",... B. Or 

 the resolution may be incomplete in various degrees, as we already 

 suspect from certain instances', in which case we may have gametes 

 .4, A A", A"'A"'\ MA'AC^.-.B, and so on. Each of "these may meet 

 a similar or a dissimilar gamete in fertilization, forming either a homo- 

 zygote or a heterozygote with its distinct properties. 



In the case of compound allelomorphs we know as yet nothing of 

 the statistical relations of the several gametes. 



Thus we have the conception 



(5) of a Compound character, borne l)y one gamete, transmitted 

 entire as a single character so long as fertilization only occurs between 

 like gametes, or is, in other words, "symmetrical;""' but if fertilization 

 take place with a dissimilar gamete (or possilily by other causes), 

 resolved into integral constituent characters, each separately trans- 

 missible. 



Next, as, by the union of the gametes ])earing the various hypallelo- 

 morphs with other such gametes, or with gametes bearing simple 

 allelomorphs, in fertilization, a number of new zygotes will be formed, 

 such as may not have been seen before in the l)reed. Thes<; will ine-.i- 

 tably be spoken of as varieties; and it is difficult not to extend the 

 idea of variation to them. To distinguish these from other ^iu•ia- 

 tions — which there must surely be — we may call them 



(6) Analytical variations in contradistinction to 



(7) Synthetical variations, occurring not })y the separation of pre- 

 existing constituent characters, but by the addition of new characters. 



Lastly, it is impossible to be presented Avith the fact that in Mende- 

 lian cas(\s the cr()ss])red produces on an average e([ual numlxu's of 

 gametes of each kind — that is to say, a synunetrical result — without 

 suspecting that this fact must correspond with some symmetrical figure 

 of distribution of those gametes in th(> cell divisions by which the}' 

 are produced. 



At the present time these are the main conceptions— though by no 

 means all — arising directly from MendePs work. The first six are all 

 more or less clearly embodied by him, though not in ever}' case devel- 

 oped in accordance with modern knowledge. The seventh is not a 

 Mendelian conception, but the facts before us justify its inclusion in 



