l38 STUDIES IN INDIAN SUGARCANES 



hairs, the width and shape of the leaves, the relations between their length and 

 width (leaf module), the erectness of the leaf tips, the length and thickness of the 

 joints and the cane module, the n.umber of joints in the cajie. the oblicjiicness 

 of the first formed shoots, the nature of the underground branching, 

 freedom of flowering and seed formation, the roughn.ess of the surface of the 

 leaves and the natuie of the serrature on their edges, and so forth. Some sixty 

 to seventy such characters are dealt with below, many of them of prime import- 

 ance, while in others the groups liang together generally but exceptions occur. 



The two new classes of indigenous canes thus differ markedly from those 

 already recognized. Thus far, we have depended in our grouping on obvious 

 resemblances by which, at a glance, varieties could be placed together, as 

 if the same cane had been grown for long periods in different tracts. The 

 members of the new classes often show no such obvious likeness, and it has 

 required a great deal of detailed study of many minute characters accordijig to a 

 prepared scheme, before we could decide on definitely placing each member in 

 its proper class. Certain varieties, such as the great, grass-green Putli Khojee, 

 are noted for the imniber of minor deviations, and there are other canes, such 

 as Shakarchynia an.d Barahi, showing obvious connections, but with sufficiently 

 important differences to prevent at present their inclusion in either class. 

 Such cases may prove of special interest later, as showing transitions between 

 the classes now dealt with and others as yet undetected, or they may be 

 simply blind variations with n.o further relationship. 



We have thus before us an interesting mass of information as to the 

 systematic value of the numerous characters which have formed the basis of 

 our cane descriptions, one of the desiderata set before us at the commencement 

 of our study. ^ In the classification of the two groups, we see that it is more 

 to the concurrence of a number of apparently unimpoitaji.t resemblances that 

 we nuist look for the proper grouping of our can.e varieties, than to characters 

 of real morphological significance, so common.ly and successfully employed 

 in the classification of plants. There is little doubt, moreover, that, to the 

 characters detailed in this paper, may be added fundamental differences in the 

 quality of the juice, fibre content, milling properties, requirements of soil and 

 water, and general hardiness and liability to disease, and there are indications 

 that such is the case. Tlie members of the Saretha class are apparently 

 hardier and less dependent on water than the Sunn.abile varieties, and also 

 have less juice but with more saccharine content. But a lack of first-hand 



^ Barber, 0. A.., Sf)me Difficulties in the Improvement of Indian Sugarcanes. Annals of 

 Applied Biology, vol I, nos. 3 and 4, Jan. 1915. p. 219. " Problem 6/' 



