A. H. Trow 247 



possible, a non -radiate lanuginosus. The necessary crosses were made 

 in 1908 and the Fi generation was raised in 1909. The result 

 appears in 



TABLE IV. 



Distribution between 



Totals 241 132 109 45 



Owing to the ease with which ci'ossing talv.es place in this type, not 

 anticipated, and of course unknown in 1908, an une.xpected result came 

 to light with the flowering of the hybrids : sotne of these xuere of RR 

 type. 



The re.'^ults were similar in each of the three cro.sses. 



In Exp. 19 (lanuginosus x praecox) witla 46 Hybrids there were 38 NR and 8 RR plants 

 In Exp. 20 (lanuginosus x erectus) ,, 39 ,, 34 NR and 5 RR ,, 



In Exp. 21 (lanuginosus x multicaulis) ,, 23 ,, 20 .NJi and 3 iJ/J ,, 



As it was important to determine whether the appearance of the RR 

 plants was due to inefficient selfing, and it was desirable for other 

 reasons to repeat the experiment, the cross lanuginosus x praecox was 

 again made in 1910 (Exp. 22) with plants very carefully isolated and 

 protected in the greenhouse, with the result that, in 1911, 31 plants 

 of the second generation were raised, cmd all of them were hybrid and 

 the thirty which Jloiuered were of NR type. In this case praecox pollen 

 completely supplanted the much more abundant pollen of lanuginosus. 

 We have therefore, however vexatious the conclusion may be, no alter- 

 native but to suppose that the mixture of RR and NR hybrids in the 

 first three experiments was due to pollination with two kinds of pollen 

 grain of R and N type respectively. It is of interest to note that these 

 RR Fi hybrids had the large ray of erectus, radiatus and that the plants 

 were of intermediate hairiness. The R pollen grains were probably 

 already on the inadequately selfed NN plants used for pollination. 



The analysis of tlie F.^ and F^ generations of the NR hybrids. In 

 these cases the analysis even of the ray character presents some diffi- 

 culty. The RR type varies considerably, — some plants have rays as 

 conspicuous as those of erectus, radiatus ; others have the rays of 

 lanuginosus. The clue to the solution of the difficulty is furnished 

 when one notices that hairiness appears to have a depressing influence 



