414 THE AMERICAN NATURALIST [Vol. LV 



The Relation of Agouti to Piebald 

 Dunn's (1920) data on the cross, agouti X piebald, in- 

 dicated a cross-over percentage, between the genes for 

 agouti and piebald, of only 46.23 ± 1.20, in a total of 783 

 young produced by a back cross of Fj animals to the 

 double recessive, non-agouti piebald. This shows a de- 

 viation of 3.77 per cent, from the normal 50 per cent, 

 value expected if the genes for agouti and piebald assort 

 independently. This deviation is barely more than three 

 times the ])i'()l)al)lt' ci-i-or. The dnta were al>o taken 

 incidentally t'roiii cros-rs in which tli.' ])riiiia rv dhjct-t 

 was to ascertain tli.' ivlal i<.!isjnp U-Uxwxx the -vnes for 

 pink-eye and picliahl. This may havr inthiciiccd the linal 

 result in sonic \\a>. ^n-h a^ h a!lc)iti.)!i hrinu' placed on 

 the discrimination of aunuti. Tiicfci'oi-c further data 



All later crosses were made with reference only to the 

 agouti and piebald factors. The original matings were 

 all for coupling, the cross being agouti self X non-agouti 

 piebald (AASSXaass). The F, animals were hetero- 

 zygous for the two factm-. ( \aS.). Such aniniaU ^Imnhl 

 form four classes of uanictcs (1) As, ( i' ) As. \'.\) aS, 

 (4) as. If the two u'cn.'s arc IiulcixMidcnt. thoc classes 

 should be e(inal in niniil..'r: ii Hnkcd, cl;,.^c. il) an. I (4) 

 should be in i'\cr^^. Whm Mich K, animaU an' hack 

 crossed to the douMr icc^hxc, nun-aunuti pichald. the 

 distribution is ohtaincd whicii is shown in Tahlc 1 oi)po- 



