15^12] Goodspeed: Nicotiana Hybrids 111 



as great as or greater than the number produced from heavy seed. 

 Conehision 1 is in accord with the generally observed facts, while 

 conclusion 2 is distinctly opposed to the report sent out by 

 those who, in agricultural practice, have matured plants grown 

 from heavy and from light weights of seed. The point of great- 

 est interest, however, lies in the fact that, if there had been 

 no careful division of the seed according to weight and if there 

 had been no especial interest in growing separately the plants 

 developed from heavy and light seed, the appearance of the 

 seedlings at the time of planting out would have led a plant- 

 breeder to choose his further experimental material largely from 

 seedlings grown from heavy seed, since they would be the most 

 vigorous and would be expected to produce more normal plants 

 than would the backward seedlings from the light seed. In the 

 light of the experimental results reported in the foregoing pages, 

 and even without further confirmation of these results, it seems 

 advisable to urge upon those experimenting in plant-breeding- 

 (1) the devoting of a greater measure of attention to the physical 

 characteristics of their pedigreed seed, (2) the making of every 

 effort to germinate all types and grades of their seed, and finally 

 (3) the bringing to development, in so far as practicable, seed- 

 lings which show all degrees of vigor and development at the 

 time when they are placed in the field. That the lightest weight 

 and even chaffy, misshapen seed will germinate under the proper 

 conditions has been shown and, of still more significance, it has 

 been demonstrated that seedlings produced from such seed, 

 though backward in appearance, may ultimately come to vig- 

 orous and normal maturity (see, in this connection, Harris, J. A., 

 1912 fo). 



From the appearance of the F^ heterozygote we seem, as 

 before stated, to be dealing with a case of Mendelian dominance 

 to the extent that the influence of the N. Tahacuni var. macro- 

 phylla parent predominates in the F^ hybrid. From the appear- 

 ance of the Fo hybrid plants we find a certain indication of 

 segregation, but that we are dealing with a case of Mendelian 

 segregation cannot, in the present instance, be demonstrated. It 

 hardly appears possible, even with the extent to which factorial 

 analysis has recently been carried and the number of IMendelian 



