1917] Goodspeed-Clausen : F^ Species Ilijhrids in Nicotiana 817 



(1881) has performed a useful service by pointing out a number of 

 these contradictions, but in many cases they have merely been 

 pointed out, no subsequent observations having been made to establish 

 the true state of affairs. In part these contradictions may be due to 

 confusion in the use of scientific names ; in pai't they apparently are 

 due to lack of consideration of the varying genetic constitutions of 

 the individuals employed, a natural consequence of the indefinite 

 ideas of heredity current before the general recognition of Mendelian 

 principles. Nevertheless some of the results appear clearly to show 

 the same general type of behavior as is reported in this paper, one 

 parent exerting by far the greater influence in the development of 

 the hybrid. 



As an example of conflicting observations concerning the characters 

 of hybrids, we may take those dealing with the hybrid rustica X 

 paniculata. Kolreuter (1761) considered that the Fj was exactly 

 intermediate between the two parents. Gartner (1849, p. 253) be- 

 lieved that paniculata exerted the greater influence in the hybrid and 

 as evidence called attention to the resemblance of "Nicot. rustico- 

 panicidato-panicnlata 5 X rustica J^" to paniculata. Obviously, how- 

 ever, in the light of modern conceptions this can hardly be considered 

 proof of the predominating influence of paniculata in the hybrid. 

 Focke (1881) from his own observations found the hybrids more like 

 rustica and considered the effect of paniculata as practically confined 

 to floral characters. Lock (1909) considered his plants of this cross 

 distinctly nearer to rustica, so much so in the seedling stage that 

 he was at one time in doubt as to whether they were actually hybrids 

 or plants from accidental self-pollination of rustica. In the adult 

 stage, however, he found them intermediate in all characters. Finally 

 East (1915) reports that the hybrid is intermediate throughout. We 

 must, therefore, look to the point of view of the investigators them- 

 selves as a factor in determining their judgment of resemblance to 

 one or the other parent. This fact is undoubtedly due in part to differ- 

 ent estimates of the comparative value of resemblances in different 

 parts of the plant. 



We may now consider a few special cases. A^. paniculata X ^^ 

 Langsdorffii is reported by Gartner (pp. 260, 469) to be very nearly 

 a complete reproduction of Langsdorffii. The resemblance of the 

 hybrid to Langsdorffii is held to be as great as that existing between 

 paniculata and plants obtained by crossing back paniculata X rustica 

 to paniculata for three or four generations. Lock (1909) and East 



