100 MendeFs Experiments 



but are sometimes so near to one of the parent characters 

 that the [corresponding] character of the other has receded 

 considerably or almost evades observation. So, for instance, 

 we see in one of the two forms of H. Auricula $ x H. 

 aurantiacum $ pure yellow disc-florets ; only the petals 

 of the marginal florets are on the outside tinged with red 

 to a scarcely noticeable degree : in the other on the contrary 

 the colour of these florets comes very near to H. aurantiacum, 

 only in the centre of the disc the orange red passes into a 

 deep golden-yellow. This difference is noteworthy, for the 

 flower-colour in Hieracium has the value of a constant 

 character. Other similar cases are to be found in the 

 leaves, the peduncles, &c. 



If the hybrids are compared with the parent types as 

 regards the sum total of their characters, then the two 

 forms of H. prcealtum $ x H. aurantiacum $ constitute 

 approximately intermediate forms which do not agree in 

 certain characters. On the contrary in H. Auricula % x ff. 

 aurantiacum $ and in H. Auricula $ x H. pratense $ we 

 see the forms widely divergent, so that one of them is 

 nearer to the one and the other to the other parental type, 

 while in the case of the last-named hybrid there is still a 

 third which is almost precisely intermediate between them. 



The conviction is then forced on us that we have here 

 only single terms in an unknown series which may be 

 formed by the direct action of the pollen of one species on 

 the egg-cells of another. 



2. With a single exception the hybrids in question 

 form seeds capable of germination. H. echioides 9 x H. 

 aurantiacum <$ may be described as fully fertile ; H. prce- 

 altum 2 x H. flagellare $ as fertile ; H. prwaltum <j> x H. 

 aurantiacum $ and H. Auricula ? x H. pratense $ as 



