158 STUDIES IN INDIAN SUGARCANES 



The thickness of the branches (canes), at 2 feet from the base, was 

 measured in each case, excepting the three es which were only strong shoots 

 five feet long, and here it was taken at the base. The figures in these three are 

 on the high side, as in these late shoots the thickness of the stem diminishes 

 slightly upwards. 



Thickness of branches of successive orders in a Pansahi clump of canes, 

 in centimetres : 



(P 



There is no doubt but that the es are, in this case, to all intents and 

 purposes, " water shoots." Two photographs of another dissection of a 

 Pansahi clump are appended (Plate III) which show this increase in thickness 

 in the later branches very clearly. 



It will be seen from the above that the late-formed shoots are branches 

 of a high order and that they are uniformly thicker than the earlier ones. It 

 is also very evident in the dissections that the joints of the earlier canes are 

 shorter than those formed later. In the special case under consideration the 

 average length of the joints in the lowest two feet of the branches of different 

 orders is as follows : — a 1", h 2", c 2", d 2"8", e 3", and this difference is especially 

 noticeable in the lowest pai-t of each cane, namely, in the first formed or basal 

 joints. The arrangement of the canes measured, in the tables of length of 

 joint in different parts of the cane, where the cane with longest basal joint is 

 placed first and those following are ones mth successively shorter basal joints, 

 is thus justified, in that, if there is any well marked division into early and 

 late canes in the twenty measured in an.y variety, it will at once become evident 

 (c/. page 30). From the study of a very large series of tables of measurements 

 of all the joints in 20 canes of each variety, always arranged in this way, it 

 may be stated, as a very general rule, that the canes with longer basal 

 joints have, on the average, fewer joints, but it by no means follows that a 

 definite division into two classes of canes, early and late, can be made at 

 crop time. We have seen that such a division is definite in members of the 

 Pansahi group. In other cases the transition from early to late cane is gradual. 

 In the varieties dealt with in this paper, of the Sunnabile and Saretha 

 classeSj all the canes were spread out at crop time in 1917, and it was found 



