SETCHELL] ABORIGINAL TOBACCOS 407 



dolph Gilmore of Lincoln, Nebraska, seed of this species which was 

 still being cultivated by a Hidatsa Indian. I have grown the descend 

 ants of the plants from this seed and in the pure line for several 

 generations and find that it still comes absolutely true to type as 

 described by Lewis and Clark and as represented by the Lewis and 

 Clark specimens. The plants very closely resemble those of the type 

 of Nicotiana Bigelovii, but the flowers are neither quite so large nor 

 so graceful. The chief difference from any of the varieties of N. 

 Bigclovii, however, is to be found in the ovary. This is constantly 

 four-celled in N. quadrivalvis, while in N. Bigclovii, it is preponderat- 

 ingly two-celled, although three celled examples are frequent in the 

 type and in the northern variety. Nicotiana quadrivalvis is not only 

 the tobacco of the Mandan, but of the Arikara and the Hidatsa In 

 dians as well. How they obtained it is not known, but it is not 

 known outside of cultivation. This latter fact, taken in connection 

 with the close resemblance to Nicotiana Bigelovii, the only essential 

 difference being the increase in the number of carpels as shown by 

 the four-celled ovary, makes it appear reasonably certain that N. 

 quadrivalvis is only a derivative from some form of A r . Bigelovii. It 

 may possibly have arisen by a single mutation or it may be a hybrid 

 derivative from a cross between N. Bigelovii and N. multivalvis. I 

 have obtained forms very close to N. quadrivalvis as descendants of 

 such a cross and such forms have appeared in the botanical garden 

 of the University of California as the result of a probably spon 

 taneous cross between the two species mentioned. It is of decided 

 interest to find a Bigelovii-derivative so far from the Bigelovii home 

 and this interest is increased by the fact that N. quadrivalvis is con 

 nected in distribution with the Californian area by the area in which 

 A r . inultivalvis, itself seemingly a Bigelovii-derivative, is found under 

 aboriginal cultivation. 



The Hidatsa tobacco, which is fairly certainly Nicotiana quadri 

 valvis, has been the subject of study by Gilbert L. Wilson. 1 He says 

 that the Hidatsa cultivate tobacco, but does not mention the species. 

 It is not used by the young men because it prevents running by 



1 Agriculture of the Hidatsa Indians, an Indian Interpretation, Univ. of 

 Minnesota Studies in the Social Sciences, no. 9, Minneapolis, 1917, pp. 121127. 



