ALBERT AND GABRIELLE HOWAHtX 77 



excluding var. frnticosa and var. lancifolia and substituting var. 

 pui'jvirea for var. mncropliyUa, 



The varieties according to Anastasia are therefore as 

 follows : — 



var. havcwensis. 



var. brasiUensis. 



var. virginica. 



var. piirpiirea. 



Recently Comes' has published a short answer to the criticisms 

 of Anastasia. 



Into the merits of this controversy it is impossible for us to 

 enter as we have not at our disposal in India the material on 

 which the rival systems are founded, and we have, therefore, been 

 unable to make use of either. Moreover, the cultivated forms 

 do not show the characters of any one of these " typical " 

 varieties, but a combination of the characteis of two or more. 

 This is referred by Comes to the hybrid nature of the cultivat- 

 ed forms, and in his lar^^e work Delle razze dei Tahacchi (which 

 is in the main a description of a large number of commercial 

 forms of tobacco from all over the world) he classifies almost all 

 the cultivated forms as hybrids between two or more varieties. 

 Thus Cimanon tobacco from Mexico is classed as a hybrid of 

 fruticosa x hrasilioisi^ x Jiavanensis x macrophyUa, China 

 tobacco as a hybrid of fruticosa x hrasiliensis x lancifolia. Such 

 a classification appears to be open to several objections. In the 

 first place, it gives no handy and workable guide in identifying 

 any particular form, and, secondly, in our present state of know- 

 ledge regarding the inheritance of characters in the species .\'. 

 tabacum it is difficult to prognosticate what products such a 

 hj^bridization of four quite different varieties would yield. It is 

 certain both from theoretical considerations, and from the 

 evidence already obtained from accidental cross-fertilization, that 

 the number of forms obtained would be numerous It does not 

 appear justifiable therefore to state that a form is a hybrid between 



' Comes, BoUrtino Tir/iiro, No. 1, Scafati, l!)(i«. 



