ALBERT AND GABRIELLE HOWARD. 3 



and Rustica. Comes considers that N. riiMica has given rise, on 

 account of the ver}^ different cUmatic conditions under which it 

 lias been cultivated from time immemorial in the new world, to 

 the followino- six varieties : — 



1. Var. ffxana. Comes. This is said to be a native of 

 Mexico and to be the first variety of tobacco introduced from 

 Florida into Portugal. The Calcuttya tobacco, cultivated in 

 India, is said by Comes to belong to this variety. 



2. Var. jamacienfiis. Comes. This variety seems to have 

 been founded on herbarium specimens only and is said to be a native 

 of Jamaica, Guatemala and Mexico. Nothing appears to be 

 known, however, as reofards the cultivation of this form for com- 

 mercial purposes. 



3. Var. asiatica, Schrank. This variety is grown in 

 Syria, Arabia, Persia and Abyssinia, and is said to be the best snuff 

 tobacco in cultivation. 



4. Var. hrasiha, Schrank. This variety is the one largely 

 cultivated in H^urope for the manufacture of snuff. 



5. Var. h> I mil is, Schrank. This variet}^ also is said to 

 be extensively cultivated for the manufacture of snuff. 



6. Var. scahra, Comes. A native of South America of 

 no commercial importance on account of its disagreeable odour. 



A careful stud}^ of the descriptions of these six varieties and 

 of the illustrations in Comes' monograph indicates that these 

 varieties are by no means well marked. Further, when it is 

 remembered that natural cross-fertilization can easily' take place 

 between the various races, it is not difficult to understand 

 that the limits of the varieties would quickly be lost. 



On attempting to refer the Indian forms to the above varie- 

 ties, difficulties were at once encountered. The majority of the 

 types in cultivation at Piisa do not resemble any one variety, but 

 possess characters, some of which agree with one of Comes' 

 varieties and the others with one or more of the remainino- 

 varieties. Furthermore, tlie Indian tj^pes form a series, and even 

 if broken up into classes, there are intermediate forms which may 

 with equal justice be placed in either of two groups. It appeared 



