i 



262 LOCK : 



Some of these cobs showed a few grains which were conspicuous i 



for the coinpletc absence (at fii-st sight) of purple pigment. 

 On closer examination a rainox'ity of these supposed non-purple , 



grains were found to show a very faint purple pigmentation, j 



although in the majoritj^ no such pigmentation could be 

 recognized.* It must be supposed that these grains were 

 produced by the action of pollen bearing an inhibiting factor 

 which is not always completely effective. 



In the case of entry No. 21 it seems probable that dark 

 p\irples are to pale purples m the ratio of 9 : 7. If this inter- 

 pretation is correct, it would appear that two separate factors 

 are rciiuircd to account for the difference between dark purple 

 and pale. The case again requires further study. 



On page 117 and the follo\\ing pages of my earUer paperf 

 are recorded the results of crosses between Black Mexican 

 Sugar Com and a number of non-purple strains. In the 

 interpretation then given I was partly misled by the facts that 

 the single example of the cross, purple male by non-purple i 



female, gave rise to exclusively jDurplc grains, whilst white X 

 purple in each case yielded a certain proportion of non-purple 

 grains. It is necessary to suppose that the inhibiting factor j 



was absent from the white individuals used in experiment 33, 

 but present in those employed in experiments 34. 35, and 37. 



'F'Ik^ offspring of the cross White Dent x Black Mexican 

 pollinalcd ivirr se are recorded in Table 13. From 18 plants 

 4,Or)2 jiuiph grams and 3,023 white gi-ains were obtained, or 

 57*27 per cent, of purple grains. This proportion would 

 apiKjar to rcjiresent a ratio of 9 : 7 (expectation 56 "25 ± 40). , 



The scries was sufficiently uniform to make it highly probable 1 



that all llic jdants were of the same constitution. , 



Tal)lc 14 shows the jjrogcny of nine similar heterozygotes 

 pollinat<'(i in bulk from plants of a supposed recessive. In 

 offjcr to aecouiit for Ihc jjroportion of purple grains produced 

 in this instance (32 6 i)er cent.) it is necessary to suppose that 

 Bouie of the n(»n-j)urp]c jiolic n bearers were pure recessives in 

 respect of both the factors V and C (expectation, if all were of 



• In one Buch cob the majority of tho pule grains were distinctly palo 

 purple. 



t Annulfl, Vol. 111., I'urt 11., November, 1906. 



I 



