F. A. Woods 305 



It is however to be noticed that the deviations from MenJcl's Law in Tables I 

 aiui II, iiaiiiely too little white, are on the sido of sii(»-gosting an anccstral influence 

 which Mendels Law does not recofrnize. This ne<,deet of ancestry Weldon con- 

 siders to be the tallacy of tlic Mendelian j)rinci|iles, and writes as follows (10), 

 p. 25-2. 



" The fundamental mistake which vitiates all work based npon Mendel's 

 method is the neglect of ancestry and the atteuipt to regard the wholu effect upon 

 offspring, produced by a particular parent, as due to the existenee in the parent of 

 particnlar structural characters ; while the contradictory results obtained by those 

 who iiave observed the otlspi-ing of parents appareutly identical in character show 

 cleai-ly cnougli that not only the parents themselves, but their race, that is their 

 ancestry, mnst be taken into aeconnt before the results of pairing theni can be 

 predicted." 



It certainly wouUl seeui that as regards the brecdiiig of mammals we cannot 

 with cur present knowledge give up the idea that ancestry has sonie influence. 



This inflnence of ancestry is contended for by Darbishire (7)* in a short article 

 in Bioinctrika in which he clearly shows the influence of ancestry in a number of 

 Grosses between the Japanese waltzing and albino mice. It seems in every one of 

 a few instances given to make a considerable difiference in the results whether his 

 albinos are from pure-bred stock or not. Perhaps further experiments may show 

 what characteristics are pure unit-characters^ , as albinism appears to be, and then 

 all results will agree closely with the expected. 



Or, on the other liand, further experiments may overthrow the reliance to be 

 placed on the few figures already collected. At any rate, the future only can 

 decide. The figures given in this article show more than anything eise, that any 

 law yet formulated, when applied to an instance similar to practical stock improve- 

 ment, gives results not remarkably in keeping with tlie observed. In this case 

 the truth seems to lie between the expected as anuounced by Mendel and that 

 formulated by Galton. 



Before closing it might be interesting to conipare the figures, 21 °/, , 

 obtained from rabbits by DR x DR with those of other observers for mice. 

 Crampe (5), p. 542, crossed grey wild mice with albinos and obtained 15 youug, 

 all resembling the wild form. The next generation gave 79 young, 22 of whom 

 were albino. S§ = 27'8 °/^. Von Guaita (8), (p. 122 Davenport's review), shows 

 under similar eonditions 14 albinos out of 44, or 31'8'7o- Cuenot (6), out of 

 270 young, 72 albino, or 26'6 % ; «■nd again 162 grey to 57 albino, or 26 7o- 

 Thus all the figures ränge between 21 and 81 '8°/,, and seem pretty close con- 

 sidering the comparatively small numbers used. 



As was stated above, my result for DR x 7i is not so close as that for 

 DR X DR; and instcad of 50 7^ R we find 40:3 7„. 



* Since ^yntil]g a secoiul article by Darbishire bas appeared. 

 t Bateson (1), p. 27, discusses " unit-eharacters." 



Biumetrika ii 89 



