292 Mr Batesons Recisions of Mendels Theory of HerecUty 



moment, as a possibility, wliat happens when the sjune white-flowered Ph. nanus 

 is crossed iu one case witli a coloured flower containing two unit coloui-s, in 

 another with a flower containing four or Hve? 



Mr Bateson seems to have perceived this difficiilty, and he has coneeived what 

 he calls a "Compound allclomorph," the sonrce of furthcr complications. He 

 conceives, if I undcrstand him, that the colour-characteis, A, and A., of a purple 

 bean flower are not, as Mendel says, independent, but that they are conibined, 

 while in the pure-bred plant or its gametes, into a System wliich behaves like 

 a Single unit, so that he would, I think, agree to write the Zygote, produced by 

 crossing the purple-flowered and the white-flowered bean together, as {Ai + A„)a, 

 where the portion enclosed in a brackot behaves like one unit character. Con- 

 cerning the behaviour of a "Compound allelomorph " when the hybrid fonns its 

 gametes, Mr Bateson has suggested a series of possibilities, eaeh involving a process 

 not ineluded in Mendel's theory, and each incompatible with the view that a simple 

 unit of colour cliaracter is comparable with " any other constant character of the 

 plant." The principal suggestious so far put forward by Mr Bateson may be 

 classified as follows: 



A. Siiggestions in harmony with Mendel's view that elementar^ unit-char acters 

 are separated during gamete formation : 



(a) The numher of elements contained in the gametes produced by a hybrid 

 need not be the same as the number contained in the pure parental gametes, and each 

 gamete of hybrid origin may contain more than one element ofthe same ki/id. 



In this case Mr Bateson supposes that a hybrid Zygote, of Constitution 

 {Ai + A., + A3 + ... etc.) a will, during gamete formation, give rise to a number of 



independent elements, A^, A„, etc and o, and these will be distributed among 



the gametes ; but he supposes that a gamete may be produced of Constitution 

 AiAi, A.2A.:,, etc. (I. p. 80, II. p. 145). This is clearly not a phenomenon com- 

 parable with anything which can be supposed to occur wlien a zygote cross-bred 

 from a grcen-wriiikled and a yellow-round pca produces gametes; for a gamete of 

 Constitution YY would have two elements determining cotyledon colour, and none 

 whatever determining shape. The Union of two such gametes would produce an 

 unthinkable result, a pca with yellow colour and without shape; to aid i>ur 

 conception of such a pea we can only ajipeal to that " residuc," the plant witii all 

 its Unit characters removed, which Mr Bateson considers to be the foundatiou 

 on which the idea of "Species" must ultimately rest ' (I. p. 28, II. p. 148). 



(6) Each gamete prodxiced by a hybrid contains one, and one only ofthe unit 

 characters transmitted to it front its pai-ents. {Froc. CambHdge Phil. Soc. xii.) 



In this case the heterozygote (A^ + A., + ...) a produces gametes of simple Con- 

 stitution Ai, A.,, ... or a, and here agaiu we have a conception which cannot be 

 applied to anything but colour without raising the difficulties pointed out in the 

 case of the prcvious hypothcsis. This secoml non-Mendelian hypothesis has so far 

 been suggested in couuection with two subsidiary suggestious : 



