48 On MeiideVs Categories 



seuse, yet it is said to have beliaved as a strictly rocessive foriii when crossod witli 

 hairy races later on. In 1892 it was crossed with normal forms ot Lijchnis diiirna, 

 and the resulting generations were held to obey Mendels laws. The hybrids of 

 the first generation wcrc all hairy ; and the offspring obtained by pairing these 

 werc in part haiiy and in part glabrous. From tiie glabrous hybrids a stable 

 glabrous variety is said to have been raised, and there is no record that these 

 glabrous " recessive " forms ever give rise to " hairy " plauts. The gametes of the 

 glabrous Li/chnis vespertina, as fixed by seleetion, are therefore said to behave in 

 this cross as if they were what writers on Mendel's work call " pure." Mr Bateson 

 rejects the view that the characters of cross-bred individuals, derived in part from 

 such "pure" parents, can be regarded as depending upon the characters of the 

 ancestors from wbich the "pure" parents are descended ; he declares that they 

 depend entirely upon tiie characters of the " pure " individuals used in making the 

 cross. It is therefore to be regretted that he has abstained from discussing an 

 expcriment in which Professor de Vries crossed this "pure" glabrous Lychnis 

 vespertina with Silene noctiflora, and obtained hybrid offspring which were indeed 

 hairy, but their hairs were of the type proper to Lychnis vespertina, and not of the 

 tj'pe of Silene noctißora. Surely we have here a clcar proof that the " dominant " 

 chiiracter, hairiness, may on the application of a suitable Stimulus be manifested 

 by the fertilised germ-cells of what is said to be a purely recessive plant ; so that 

 the theory of pure parental gametes, on which Mr Bateson lays ."^uch stress, is 

 shown to be inade(iuate for this ca.se, and a theory of inheritance, with rever.sion to 

 particular ancestors, is indicated as likely to express the facts of " Mendelian " 

 cases also. 



As has already been said, Professor de Vries crossed the glabrous L. vespertina 

 here doscribcd and normal hair}' L. diurna, always apparontly using L. diurna ? 

 X L. vespertina ^ ; the first generation of hybrids contained only hairy individuals. 

 The secoud generation is not very fuUy described ; in the Erfelijke Monstrositeiten 

 (p. 72) we are told that about § of the individuals were hairy, and ^ glabrous ; in 

 the Comjdes Rendus, March 1900, we are toid that 28 per ccnt. were glabrous; and 

 this Statement is repeated in subsequent accounts. 



Although the actual numbers of individuals are not given, so that the probable 

 errors of these results cannot be calculated, it is clear fi-om the adoption of the 

 round numbers i^ and § that the imprcssion produced was not that of \ recessive 

 and £ dominant or dominant-hybrid individuals, as it should have been on 

 Mendel's hypothesis. Several hundred of individuals are said to have been 

 observed. The odds against a deviation so largely exceeding Mendel's rcsult with 

 500 individuals would be about 17 to 1. The glabrous variety produced in these 

 experiments appears to have furnished some, at lea-st, of the glabrous, rcd-tlowered 

 forms used by Mr Bateson and Miss Saunders, aud in their hands it is said to have 

 given results in good accord with Mcndd's law ; they pass over the deviations 

 from Mendel's law, observed during its earlier history, without notice. 



