G. A. BARBER ill 



The rest are in fairly close agreement with one another. It is interesting to 

 note that, in the Saretha group, the division into the Katha and Mesangan 

 sections receives support, in that Katha, Saretha and Chih develop much 

 earlier than Ganda Cheni, Khari, and Hiillu Kabbu. A similar cleavage now 

 shows itself in the Sunnabile group, although this was only suspected when the 

 varieties were examined in Memoir III. Dhaulu, Bansa, and Mojorah 

 develop earlier and branch more copiously than Dhor, Naanal and Sunnabile. 

 The position of Mojorah, the thickest in the group and nearest in several 

 respects to thick canes, was unexpected, but it is worth while noting that, 

 both in the Saretha and Sunnabile subdivisions, the cleavage indicated by the 

 dissections is according to geographical regions, the earlier matming, more 

 richly branching forms being clustered along the Himalayas, while the tardier 

 varieties are found in the Peninsula. In the Nargori group, which is generally 

 marked by the homogeneity of its members, Keimli and Ketari differ some- 

 what from the rest. But in spite of these irregularities, the general result 

 of the dissecting work has been abundantly to justify the general lines of 

 classification adopted, and, as shown in the study of the Saretha and Sunnabile 

 groups in Memoir III, the branching system yields a character of systematic 

 value. It must be remembered that only a few varieties in each group have 

 been studied, and these remarks on classification must therefore be regarded 

 as suggestive rather than otherwise ; but it is worth recording that subdivisions, 

 such as are obvious in the Saretha group, may also be expected to occur in the 

 other main classes, and it is hoped that the apparent uniformity of the various 

 groups may break down on further study, for this is to be expected in any 

 natural system of classification. 



The form of the diagram, recording the dissections of plants in full grown 

 clumps, is similar to that given for the Pansahi plant figured in Memoir III 

 (p. 157), and the character of the lines used for the branches of different orders 

 is the same as in that figure. Certain conventions have been introduced, 

 which may be summarized as follows : — 



(1) Sleeping buds are altogether omitted as having nothing to do with 

 active branching. Only those much swollen or bursting have been included 

 and their relative size is indicated. Large, white, clawed buds are, it is pre- 

 sumed, still under ground, and the dividing character between them and the 

 small shoots is the presence or absence of any green at the tips. 



(2) Dead buds are indicated by a short line with a cross fine at the end. 

 The same cross line shows dead shoots or canes. Most of the diagrams have no 

 distinction between resting and burst buds, which have died. 



