am SS ‘en tei ag Wines ite See 
1913] EAST—CROSSES OF NICOTIANA 179 
This designation was manifestly incorrect, as the plant was exactly 
like SANDER’s figure of N. forgetiana in the Botanical Magazine 
(No. 8006). As it had been collected only a few years, I took seed 
from one of the capsules and planted it. It grew and again pro- 
duced plants like N. forgetiana. Miss Day, the librarian of the 
Gray Herbarium, then looked up the correspondence regarding 
the specimen and found that it was evidently a garden specimen 
grown by a Californian botanist, since deceased, from seed fur- 
nished by Sander & Sons and called N. Sandarae. The plants have 
not the mixed colors and the variability of the specimens now sold 
as Sandarae hybrids, but are constant in their characters and are 
identical with Nicotiana forgetiana. I have come to the conclusion, 
therefore, that I have obtained (as Lock? probably did) seeds of 
the real NV. forgetiana that had been mixed with the Sandarae’ 
hybrids by Sander & Sons. 
In view of the fact that NV. alata grandiflora and N. forgetiana do 
not differ essentially in their foliage and habit of growth, but only 
in flower size and color—the one being white, the other red—per- 
haps one should not call them two species. I hold no brief either 
way. Isimply accept the taxonomic ruling. At least, there existed 
here two strains very different from each other and very constant 
in their characters. Both were self-fertile, and in fact were usu- 
ally self-pollinated naturally. They were crossed. There was no 
trouble about this, as every cross attempted was successful, and 
the capsules were filled with seeds. 
This, then, seemed to be an excellent opportunity for studying 
size inheritance: two strains, uniform in pure lines, one with a 
corolla three times the length of the other, could be crossed easily. 
All was not plain sailing, however, for the plants of the F, genera- 
tion (pl. VII) were absolutely self-sterile. This fact would have cut 
off the experiment in the flower of a promising youth but for the 
further fact that each plant was perfectly cross-fertile with every 
other plant. It did indeed reduce my interest in the inheritance 
of corolla size, for it precluded the study of an F; generation, but 
this was offset by the more fascinating problem of self-sterility. 
2 Ann. Roy. Bot. Gard. Peradeniya 4:195-227. 1909. 
3 The Sandarae hybrids were supposed to have been produced by the cross V. 
forgetianaXN. alata grandiflora. 
