30 JOURNAL OF THE ROYAL IIORTICI LTURAL SOOIF/FY. 



tall ' Black Knight ' with ' Pink Ciipid.' In this cnso, however, three 

 pairs of characters were concerned in the cross, and consequently, on 

 the average, only three plants out of sixty-four were expected to he 

 ' Black Kniglit Cupid/ and only one of these sliould hreed true. 

 The actual nunihers were as follows: — 



In the first experiment, out of 244 plants there were ten * Black 

 Knight Cupids,' the Mendelian expectation nearest in whole numbers 

 being eleven. In the second experiment, out of 166 plants there were 

 seven * Black Knight Cupids,' the expectation being eight. In the third 

 experiment, out of 165 plants there were only three * Black Knight 

 Cupids ' instead of the expected eight.* 



Adding the results of the three experiments together, there were 

 out of 575 plants twenty * Black Knight Cupids,' where twenty-seven 

 was the Mendelian expectation.! 



In the next generation (F,) some of the ' Black Knight Cupids ' bred 

 quite true, while others did not. Curiously enough, one plant gave a 

 mixture of ' Black Knight Cupids ' and Wild Purple Talh, which was 

 probably due to the intervention of bees. 

 The following diagram (fig. 16) will illustrate the above experiment : — 



Sweet Pea 



Tall Cupid 



PuHPLE RtD 



DarkWing Light Wino 



Tall 

 Purple 

 l i c h t w i n g 



Fx I 1 1 — i ' ' T — I 1 » 



TPL TPD TRL CPL TRD CRD CRL CRD 



27 9 9 9 _ 3 3 -*.^.^^__3^ 1 



Fig. 16.— Showing the origin of the ' Black Knight Cupid ' Sweet Pea 



BY re-combination FOLLOWING MeNDELIAN SEGREGATION. 



T = tall ; C = Cupid ; P = purple ; K = red ; L = light wing ; D = dark wing. 



T, P, and L are higher factors; while 0, R, and D are the cor- 

 responding lower factors. 



* In this experiment there were ten ' Cupids ' which somehow failed to flower, 

 owing probably to their being overcrowded. 



t In these experiments there was throughout a fairly uniform excess of tails, 

 purples, light wings, and long pollens, with a corresponding deficiency of Cupids, 

 reds, dark wings, and round pollens, the cause of which is at present obscure. 

 On the whole the numbers were nearer 4 : 1 than the expected 3:1. There was 

 also a partial coupling of purple colour with long pollen and red colour with 

 round pollen, but all the other characters were inherited independently. 



