128 PLANT HYBRIDIZATION BEFORE MENDEL 



give products constantly sterile." (p. 228.) On the other hand, 

 he further comments (p. 254) : 



"Crosses between two races or two varieties give, on the contrary, as 

 Kolreuter has established, and as all those have recognized who^ have 

 followed in his footsteps, products as fertile as legitimate species." 



Godron's point of view as to the value attaching to hybrid 

 studies, is shown by his remark : 



"This fecundity then, equal to that of the parents, characterizes crosses 

 (metis) and offers us a criterion to distinguish what is a race or a variety 

 from that which is a species." (p. 255.) 



As to the fertility of hybrids and the perpetuity of their char- 

 acters, he cites especially the case of Aegilops tnticoides polli- 

 nated with pollen of wheat, and giving as a result Aegilops spel- 

 taeformis, which, he says "at first fertile to a mediocre degree, 

 like all hybrids of the second generation, produces, in the follow- 

 ing years, as many seeds as any Aegilops or Triticum known, 

 (p. 272.) The fact that the fecundity of the hybrids does not 

 always bear a relation to the facility with which the cross is 

 effected in the first place, is illustrated by Godron from Verbas- 

 cum crosses, especially Verbascum austriaco-nigrum X phoem- 

 ceum. (p. 283.) This sterility he recognizes as being due to one 

 or several possible operating causes : The complete absence of 

 pollen, defective pollen — deformed, etc. — or physiological steril- 

 ity, as in the case of Antirrhinum majus, A Barrelieri^ which, al- 

 though having an abundance of pollen, apparently completely 

 normal, yet remained entirely infertile. 



Godron again comments on the very great vegetative develop- 

 ment in hybrids of Verbascum : 



"The numerous branches, and the immense quantity of flowers which 

 arise on these branches, would they not exhaust the vegetable juices at 

 the expense of the organs of reproduction ^" (p. 287.) 



With regard to the question (p. 289) whether hybrids, self- 

 fertilized, sometimes retain their characters unchanged for sev- 

 eral generations, and thus become the type of constant races, or 

 whether, on the contrary, they always return to the forms of 

 one of their parents after several generations, Godron gives his 

 case of Linaria hybrids, stating that : 



"these hybrid forms may become very fertile, and a certain number of 

 individuals return, after the second generation, to the one and the other 



