C, A. BARBER 141 



seedlings, the mode oi' brancliing, etc., and we have utilized these resemblances 

 to Saccharum spontaneum as strengthening our conviction that the Saretha 

 series is the more primitive. But there is also some reason for supposing that 

 the more highly developed Sunnabile varieties are also traceable to Saccharam 

 spontaneum. . Among the characters in which the Sunnabile and Saretha groups 

 differ are the bending of the leaf tip, the erectness of the young (germinating) 

 shoots, the colour of the stem, the ligule and its setse, spines oj\ the backs of 

 the leaf sheaths, scar band and scar line, and so on. Now a more or less 

 imperfect study of the seedlings of Saccharum spontaneum raised at Coimbatore 

 has also shown these very differences among themselves ; there are white and 

 brown stemmed forms, the Ugules vary and the setie are occasionally long, 

 there are spiny and spineless forms, and there are variations hi the scar band 

 and scar hne. When we consider the extrenaely primitive n.ature of Dhauh, 

 and its obvious resemblance to Katha in many particulars, there seems to be 

 some justificatioii for the expectation that a closer study will increase the 

 number of these resemblances in some seedlin.gs of Saccharum spontaneum to 

 the Sunnabile series. There is thus some reason for tracing both of the primi- 

 tive Punjab canes to a common ancestor, Katha from a red form and Dhuulu 

 from a white, the diff'erences in whose seedlings have been fixed and transmitted 

 by the vegetative method of propagation. 



The other groups of indigenous Indian canes have not as yet been subjected 

 to the detailed comparative study accorded to the Saretha and Sunnabile classes. 

 It is by no means impossible that, when this is done, similar resemblances 

 may be detected. Judging from the single character, the mode of branching, 

 it would seem that the Mungo and Pansahi groups are the more primitive, and 

 that they fall into line with the Saretha class, while the Nargori group, as far 

 as it has been studied, shows a formula more comparable with the Sunnabile 

 series, a distinct advance towards the thick canes. Khelia and its allies are 

 yet further developed, being intermediate in this respect between the Sunnabile 

 and Nargori groups an.d the thick canes. We thus see that the study of merely 

 one character may thrown hght on the various stages of development of Indian 

 canes, and it is justifiable to assume that a comparative study of man,y charac- 

 ters will make the path plainer. The apparent absence of differences such 

 as we have collected in this paper hi the members of the Pans-ahi, Nargori, and 

 Mungo groups may be due to their being more recent offshoots from the parent 

 stem. Lastly, there are certain primitive canes growing in the Central Pro- 

 vinces which should be compared with Saccharum spontaneum, e.g., Barahi 

 and Katai. The latter might lead to a connection with the Nargori group. 

 So also Shakarchynia and Kheri in Bengal, which appear to be primitive forms, 



