Wilson — Unsound Mendelian Developments. 415 



While the presence and absence theory, being unsound, must lead to 

 erroneous conclusions, some of the work it has helped to produce is workably 

 sound. This happens in cases where no real recessive is identified to raise 

 confusion with the unidentified " absence." The tiieory originated in the error 

 of taking characters due to several causes to be due to one ; and it fails 

 in assuming a factor's absence to be its own recessive, with the result that, 

 when a real recessive is identified as active, a factor just one too many for the 

 case to hold has to be introduced, as we saw in the case of the rabbit colours. 

 We shall see this if we analyse by the Mendelian method a case which 

 has been brought to a correct conclusion by the presence and absence method, 

 and incidentally we shall see the power of the Mendelian formulae in 

 analysis. We shall take the case of mouse-colours dealt with by Cuenot and 

 Miss Durham. 



Miss Durham's first experiment, in which two pairs of characters are 

 concerned, was with agouti and chocolate mice. The second crosses were 

 9 agouti : 3 cinnamon agouti : 3 black : 1 chocolate.' Write down these 

 groups with the non-committal scheme below: — 



In Miss Durham's second experiment, the second crosses from black and 

 silver-fawn were 9 black : 3 blue : 3 chocolate : 1 silver-fawn.^ From the 

 fii'st experiment we know black to consist of ^^Q and chocolate of pq. As 

 to the other characters in the case, we can only write down non-committal 

 symbols. Thus the provisional scheme becomes 



The new characters may or may not be the same as the previous ones. &• 

 is obviously the same as Qq, or a new pair in which S concurs with Q and s 

 with q. In that ease the two pairs could not be separated, and we therefore 

 take Qq as representing both Qq and &, which are either the same or two 

 inseparable pairs. The pair Rr is obviously new, since it can concur with no 

 other pair already present. It is just possible for R to be the same as ^j, 



' Evolution Committee Report, iv, p. 42. ^ Ibid. 



SOIF.NT. PKOC. E.D.S., VOL. XHI., NO. XXVII. 3 Q 



