No. 609] THE THEORY OF THE GEXE 



521 



(A) Assumption of (knwtu r,nhn. A>hitnnn 

 Ithas'been said that })y assiuniiiji- enough a'ciictic factors 

 you can explain anything. Tliis is ti ue ; and it is 

 the p:reatest danger of the factorial procedure. If, for 

 example, whenever one fails to account for a result lie 

 introduces another factor to take care of what he (-iu\ not 

 explain he is not proving anylhlnu' cxcc])! tlmt he i- in- 

 genious or only naive. To make uoud llic i nt lo. hid ion 

 of another gene in Mendelian work, its presence must he 

 established by the same kind of evidence as that on whicli 

 the existence of the original factors was established. For 

 exam))le. Bridges found that after eosin eye color had 

 IxHMi ciosscd to a certain red-eyed stock, there appeared 

 in ]at<M- ucnernlion^ a new cla^s of eye color (Cream II) 

 that was far liuliter than eosin. He isolated this new 

 chaiacter and showed that the difference between it and 

 eosin was due to a specific gene that in iniicritinicc lit'hax v's 

 like other genes, although its action is not nppatciit on 

 the normal red eye, but is evident on the cnc cohir co-in. 

 IFcrc, tlien. thron'oh eN])eriinental tests. th.'"ncliial dcnion- 

 ^ii.itlnu mad.' that tlic chan-r in color of tlic cumii 



ua^ dm. tn another ucne lii<i.ien in tho noniial M<M-k. 



il-i) S/ahiliti, nt (inns rrrsus I u st a I „} i t „ n/ ilnnnrt, r.< 

 It 1 ,1^ keen ..l.iori.-d ti at it i^ unica-uiahlr lo a^Hinic 

 tliat uvn(- arc rclativciv .tabic. Tlii^ ohjcctiun l.a^cd 

 lai-cly ,.n the fact that character.^ arc n.)toriou.lv fluctu- 

 aliiu. and ^in.M- .-liai .u-tei ^ lonn tnc l.a^i^ ..l our mmicr- 

 iral .lata frc-m which tlu' i<lca of the -cue i^ .h'rivcd. it is 

 suppoMMl that -eiic> to,) mn^t l)e variahlc. Tlii^ i> bv all 

 od<ls the mo-^t comnuMi critici>,n that lia> keen broudit 

 auaiii-1 the i,]ea of u^nrtie fa. to,-, and lli.- Mi...t .lifticult 

 one t.. .k.pioxe. TI,.,.. ai.'lix..;m.vN.M.. h..u ..x ..r. to this 



In the lii-t pla.-c it ha> ht.en >iiown in a number of 

 install.'... that th.' \ arial.ility of the .•hara.-t(.r is due to 



