406 BOTANICAL GAZETTE [MAY 
possible correlations between seed characters and plant characters, and on the 
inheritance of certain quantitative character complexes in crosses between 
various Nicotiana species and varieties. 
In that part of the contribution concerned with somatic correlations 
the author deals with a cross between the varieties virginica. and macrophylla 
of the species Nicotiana tabacum. F; seeds were divided arbitrarily into the 
classes light, medium, and heavy. The light and medium seeds germinated 
more quickly than the heavy seeds, and plants resulting from the former 
matured more quickly than those from the latter. It does not appear, however, 
that the general belief that heavy seed gives more vigorous plants than light 
seed is incorrect. oreover, the total germination of heavy seed was higher 
than that of the other two classes. 
The author continually speaks of the dominance of one plant over another, 
a mediaeval mode of expression that makes it impossible to draw any conclu- 
sions from his observations on his F, generation. Likewise he finds that the 
heavy seed gave, in F,, 39 per cent of “dominants” (resembling macrophylla), 
r cent of intermediates, and 9 per cent of “recessives’’ (resembling 
ing results it is not surprising that he invents a theory for their interpretation 
that will no doubt be very interesting to cytologists, for in it he assumes that 
the “tube nucleus” of the pollen grain unites with the fusion “endosperm” 
nucleus of the embryo sac. He assumes that there are two “‘determiners,” 
one functioning to produce virginica characters and the other macrophylla 
characters. The generative nucleus, he says, may bear either, and the “tube 
nucleus” may bear either. The same alternatives are assumed for the egg 
nucleus and the fusion ‘“endosperm’”’ nucleus. Then, simply by having a 
macrophylla generative nucleus unite with a macrophylla egg nucleus, and a 
macrophylia ‘tube nucleus”’ unite with a macrophylla fusion nucleus, he gets a 
heavy seed having macrophylla characters. Quod erat demonstrandum. Chari- 
tably granting that the words ‘“‘tube nucleus” were slips of the pen, there is no 
excuse for founding a theory that the two male nuclei carry different ‘“deter- 
miners” upon unsupported data of this character. 
In the second part of his work, the author has studied the degree of domi- 
nance and the variation in size of corolla diameter in the parents and F, genera- 
tion of crosses between varieties of Nicotiana acuminata. The corolla breadth 
of the F, generation was found to be the arithmetical mean of the two parents. 
The fluctuation in corolla breadth, both in individual plants and in the popula- 
tion as a whole, was greater in F, than in the parents. These conclusions, at 
least as to the degree of variability of the F, generation, are at variance with 
the results of several careful investigators (the reviewer can count twelve 
such offhand), but it is impossible to criticize GoopsprEp’s data, for he does 
not give them in the form of frequency distributions. He simply reports 
maximum and minimum measurements, which may or may not mean anything. 
» 
