{•.. N. COLLINS ANH J. II. KLMI'TON. - IMIEIIIÏANCI': OF WWV K.MKtSl'EUM. :>h-) 



ther can Ihe piirily of types be assuiiu'd willi sucli conliileiice siiice Ihe same or 

 similar colors tliat appear in (he Chinesc types are coniinonin American varieties. 



In ail our crosses between Chinese, willi coloreil aleurone and white varie- 

 ties, the color bas been dominant in Ihe first génération, but the beliavior in 

 subséquent générations bas been l'ound to be very irref,''ular. Tbere is, iiowcver, 

 a very dcfmite relation ])et\veen tlie appearances ol' lliewaxy endosperm and the 

 color ot' Ihe aleurone. In every case \\ hère the parents difl'ered both in texture 

 of the endosperm and Ihe culor ol' llic aleurone, there was found lo lie a corré- 

 lation or cohésion beLweeu Ihe characlers possessed by the individual parents. 



If Ihere was no corrélation we should expect to find Ihe same percenlage 

 of waxy seeds among the colored seed as among the white. This, it will be ob- 

 served. is mit Ihe case. As a measure of the degree of the corrélation, the 

 coefficient of corrélation, or association, bas been caiculated by Yule's for- 

 mula'. By this met hod, when any class is lacking, as for cxample, if none of the 

 colored seeds were waxy, the corrélation is considered perfect and would be 

 represented liy I. If the percentage of waxy seeds in the colored and white 

 seeds is equal, Ihe corrélation would be 0. The décimais bcl\ve<'n and 1 

 represcnt the inlermediale degrees. 



In the following tables Ihe four classes as Ihey occurred in Ihe dilTcrent ears 

 are shown, together with the coefficient of corrélation. The coefficient of cor- 

 rélation ranges from TiSS to 89.j in the individual ears. The probable error of 

 the coefficient in ail cases is in the second décimal place and in Ihe car w ilh Ihe 

 lowest corrélation is less Ihan one sevcnlh Ihe coefficient. 



Detaih of the Ihjhr'nh Hhinriwj Corrrlalion Brdiu'eii Color o/ 

 Alcuronr ami l'Iiarai-lvi' of Ihi' fùhlot^jifnii . 

 CroSf'C» of n'/iilr l'/iiiir^i' irill, f, dorai Atorriron Vririclies. 



In 1009 a plant of AT, a semi-slarchy variety wilh a Ihin horny layer and 

 colored aleurone grown by Ihc llopi Indians of Arizona, was poUinated by DU, 

 the number assigned to the white strain of Chinese with waxy endosperm. The 

 resulting car (Dh 15) had ail the seeds colored and with horny endosperm. Expe- 

 rimenls had shown the colored aleurone in A7 lo be récessive lo the absence of 

 color in the same variety and in several olhers. 



l'ive plants grown from Ihe Dhla ear in 1910 were self poUinaled (See 

 Table 111 1. The percentage of waxy seeds for the individuel ears varied from 

 20,9 to .'0,1. The percentage of white seeds varied from 27,8 to 48,7, In every 

 ear there was a corrélation between the colored aleurone and Ihe horny endo- 

 sperm, 



TAliLK III 

 llopi Hhuhx, WInIr (Jinictr Ilh V.y Srtf-potliiiiilrd . 

 l'KD. Nci. NO. niiin; m», \miitl. ni. cm.. mi. i.ul. 



"F [L\K. NU. ~EI.MS. IIOIINÏ. \\ A\K IIIUOÏ. WAXV. (OUR tl, ITIUN. 



151 ."itjS 77 111 545 5.5 .7(i'2 



152 IN.-. i'I i'.\ 112 20 .701 



Ijô 081 149 ll'.l 2,=i7 .jO ,o7!) 



154 601 SO lOll 417 JS .895 



155 âiO 172 91 2.-,S 59 ..555 



lOTAi.. . . 2.555 500 450 1.5liU 2-i!S .741 



1. Yn.E l'G.-U.). On (lie Asïocialioii of .Ulriliulcs in Slalisllcs. Vhil. Trans. Iloy. Soc. vol. 19i. 

 1900, pp. 257-319. 



IV" CONFÉn. INTERN. HE CLNtTIQrE. 23 



