7G STUDIES IN INDIAN TOBACCOS. 



over the stigma that the latter is never visible. The filaments of 

 N. tahacvm are much longer and the flowers are larger than in the 

 case of the yellow-flowered tobaccos. In consequence the anthers 

 do not fill up the throat so completely and are not in such close 

 contact with and do not form a rincv round the stiofma as in tlie 

 case N. rusfica. Sometimes the anthers are on one side of the stigma 

 only ; in other cases some are on one side and some on the other. 

 These circumstances affect pollination considerably, and as the 

 flowers lie more or less horizontally, a good deal of pollen is wast- 

 ed and often falls on the inside of the throat of the corolla. The 

 types in which the stigma is above the stamens never set seed 

 when bagged and have to be selfed. 



3. Scheme of classification. 



A comprehensive phylogenetic and S3\stematic stud}^ of the 

 species N. tahacnm was made by Comes" who collected a large 

 number of the different forms from all parts of the world. From 

 the results of this study he divided the species into six varieties, 

 the distinction between them depending very largely on the shape 

 of the leaves. 



(1) var. fruticosa Hook. 



(2) var. Jancifolia (W.) Comes. 



(3) var. vivgmica (Agdh.) Comes. 



(4) var. hi'asiliensis Comes. 



(5) var. havanensis (Lag.) Conies. 



(6) var. macrophylla Schrank. 



Of these he states that (3) and (4) are no longer to 

 be found in a pure state, but only as hybrids with other 

 varieties. 



This subdivision of the species has been vigorously assailed 

 by Anastasia^ who reduces the number of varieties to four, 



* Comes, Monofiraplin' dn grnrc Mrotiana, Naple?, lSi»l) .and I)e/tr ra:zc del Tahacrln, 

 Naples, 1!K»5. 



2 Anastasia, Lr ranrta tijiirhr dt-lla Mcotiana tnhanim, L., Scafati, I'.tO?. 



