A. H. Trow 257 



impossible to arrange the forms in their respective groups with strict 

 accuracy. 



The observations summarized in Table VII suflSce to demonstrate 

 that hairiness and its opposite are transmissible characters, and there- 



fore probably represented in the constitution of these plants by at least 

 one pair of factors. We shall therefore make use of the signs HH, Hh, 

 and hh to represent the three types of constitution. Glabrous forms 

 are hh. We shall find that the variability of the hairy character is due 

 (1) to the influence of other factors, and (2) to a second pair of hair 

 factors ^ 



Exps. 19, 20 and 21. In 1908 lanuginosus was crossed hy pra£Cox, 

 erectus and multicaulis. Lanuginosus is hairy {HH) and radiate (RR). 

 The other three types are each glabrous (hh) and non-radiate {rr). 

 The objects in view were: — (1) to secure the non-radiate form of lanii- 

 ginosus, (2) to secure hairy varieties of praecox, erectus and multicaulis, 

 and to determine whether the hair factors could be transposed from 

 type to type with the same facility as the ray factors, and (3) to note 

 the eflfect on each other of the factors for hair and rays. The results 

 were unexpected. At the close of the examination even of the F^ 

 generations of these three crosses, it was still somewhat doubtful whether 

 either of the first two objects had been secured : the third however had 

 been investigated with at least considerable success. 



The Fi hybrids in each of these three crosses are of intermediate 

 hairiness ; in the F^ generation segregation takes place, and at least 

 three types of hairiness are noticeable. Taking the hairiness of the 

 Fi hybrid as the standard, we find (1) some plants conforming to this, 



> See notes to the illustrations for evidence of a third pair of hair factors. 



