246 BOTANICAL GAZETTE [MARCH 
four-celled capsule. Unfortunately the cross between the normal 
two-celled N. Bigelovit and N. quadrivalvis has been lost. It is 
interesting from the standpoint of the transmission of that char- 
acter, and will be remade. 
The similarity of the two plants and the fact that they give a 
cross that is fertile in the F, generation is sufficient evidence to 
convince me that N. quadrivalvis Pursh is really N. Bigelovit var. 
quadrivalvis. There is further evidence in the fact that NV. Bigelovit 
has produced a guadrivalvis variety while under observation. 
Fic. 3.—At left, Nicotiana Bigelovii var. quadrivalvis; at right, N. quadrivalvis 
Pursh; in center, F; generation of reciprocal crosses. 
Several plants from the different selections of N. Bigelovit with 
a single capsule having three cells were observed. Seeds from these 
plants were selected with the object of producing a race having 
three-celled capsules. Selection had absolutely no effect. Among 
the progeny an occasional three-celled capsule was found, but the 
percentage could not be increased. In this strain of N. Bigelovii 
there was evidently no ability to transmit the three-celled char- 
acter. It simply gave an occasional zygotic variation of this kind, 
just as do many other species of Nicotiana. 
On the other hand, several other plants, typically NV. Bigelovii 
in size, produced several three-celled capsules. It is quite probable 
that they were all progeny of one plant of the preceding generation. 
One of these plants was selfed and the resulting seed planted on 
