W. F. 11. Wkldon 289 



separate what he considers cssential iu thcm tidiii wliat is not. Tlic |ilicnomenon 

 of doniiiianee, thongh ahvays observed hy Mendel, is not essential; and in assertinji; 

 this proposition Mr Batesori does not hesitate to accuse lii.s niaster of something 

 very like a gross breach of good faith. Mendel's description of the result obtained 

 by Crossing peas is translated by Mr Bateson in these words : " In the case of each 

 " of the seven crosses the hybrid character resembles that of one of the parcntal 

 " forms so chjsely that the other either escapes Observation coinpletely or cannot 

 "be detected with eertainty" (I. p. 49). Tliis is a clear and uncipuvocal Statement 

 of an observed experiinental result, that in each of seven crosses dominance was 

 observed. No honest observer eonhl have formulated this stateuient if he had 

 observed any large inimlier of e.\eeptions, and it secnis to me that no one cau read 

 Jlendel's paper witliout believing hirn to be absohitely honest; yet Mr Bateson 

 thinivs it faii- to suppose that the bchaviour of 12 out of 34 varieties of peas, 

 cnltivated by Mendel, was not recordccl in his final statement because amongst 

 other drawbacks they exhibited " defective and irregulär dominance" (I. p. 137). 



By an analysis of this remarkable kind we are led to the statement that what 

 is essential in Mendels work is no ascertained fact at all, but Mr Bateson's peculiar 

 Interpretation of the hypothesis concerning the nature of germ-cells, whiuh he calls 

 "the great fact of gametie purity " (I. p. 117, II. ]>. 12). 



3. Atavisin. in hybrids. — By rejecting dominance as " unessential," Mr Bate.son 

 is able to apply the theory of •' pure " gamete formation to a number of cases in 

 which the ofTspring of hybrids exhibit real or apparent reversion to the characters 

 of the pure pai'ental races, although nothing like dominance is seen iu the hybrids 

 theniselves. But here he meets with a dittieulty. In many of the newiy-included 

 cases the hybrids exhibit characters which suggest reversion to remote ancestors ; 

 for example, wheu von Guaita crossed pink-eyed albino mice with dark-eyed, black 

 and white waltzing mice, he says that the hybrid young always resembled a wild 

 mouse in size, colour, and wildncss ; here we have at least au apparent reversion 

 to some characters of the common ancestors of both races. Again, Mr Bateson 

 finds that there are several distinct races of Sweet Pea, which give rise to similar 

 hybrids when crossed ; and of these hybrids he says " it is characteristic of such 

 "forms that they may reproduce in appearance some putative ancestor" (II. p. 144). 

 It is not thought necessary to include facts like these, which surely depend upon 

 the Constitution of the germ-cells as much as any other phenomena of inheritance, 

 in a theory of the gametes concerned. The apparent atavism is said to be peculiar 

 to heterozygotes, and is so dismissed ; its bearing on the phenomena of cross- 

 breeding as a wholc, and especially on " the great fact of gametie purity," being 

 neglected. 



We have here a second addition to Mendel's Statements — the manifestation of 

 a spurious atavism as a property of hybrids, and the doctrine that several different 

 gametie combinations may produce identical individuals. 



So far as I can discover, from an attempt to coUate Mr Bateson's various State- 

 ments, the only thiug which can be predicated of the result of cro.ssing pure races 

 Biometrika ii 37 



