o Ilhodora [January 



of V. affinisXsepicntriojialis have been under cultivation since July, 

 1904, and they all look just alike. About twenty plants of T. cucullafa 

 Xfimbriatula, from three different States show no appreciable differ- 

 ence. Also the infertility of these hybrids is most noticeable; rarely 

 are more than 30 per cent of the ovules fertilized, usually about 10 

 per cent, occasionally none. Focke^ has noted the marked infer- 

 tility of the European hybrids of Viola. Ivastly, the rank growth of 

 violet hybrids is most pronounced. In many cases I have the hybrid 

 growing side by side with the parent species, and always the robust- 

 ness of the hybrid is strikingly apparent. I have frequently made 

 eight or ten am))le specimens out of one plant. On one individual 

 of V. affiiiisXsororia I counted last May 148 large petaliferous flowers. 

 But when crosses are made between doubffully distinci species, or 

 between races, or between a species and its variety, there is commonly 

 a marked departure from the first three of the laws above stated, viz. — 



1. The offspring of such crosses have not intermediate characters, but 



various recombinations of the unlike characters of the parent forms. 



2. The individual plants are consequently often dissimilar, some reverting 



to one or the other of the parent forms, others presenting all sorts 

 of compromises, — a phenomenon known to breeders as "sporting." 



3. There is no impairment of fertility either in the first or in subsequent 



generations. 



To this class belong the so-called Mendelian hybrids,— the despair 

 of the systematist, but tiie vantage ground of the breeder of new and 

 useful "varieties." 



The latest diseussion of tliis subject is to be found in the recently 

 published volume of DeVries. This author makes a sharp distinc- 

 tion between a specific and a varietal character. The latter he regards 

 as but the loss or latency of a quality, which is positive and dominant 

 in the typical form of the species. When interbreeding takes place, 

 he holds that specific characters are affected according to the first set 

 of niles above cited; but varietal characters according to the second 

 set. I can make but the briefest allusion to this novel ex])hmation 

 of the phenomena of hybridism, and my excuse for so doing is that it 

 seems to find illustration in the behavior of certain crosses between 



' L. c, p. 477. 



