171 



StUDIES IN lifDIAN StJGARCANfiS 



thickness, cacli cane is measured in three places, at the base, in the middle and 

 at the highest mature joint, and these measurements are made, by means 

 of calipers, in two directions, median or in the plajie of the bud, and lateral, 

 or afc right angles to it. 



Of these figures the first is always lateral, and it is seen that ovalness is a 

 constant character, the median, thickness being always greater than the lateral. 

 There is little difference bevveen the two groups in ovaluess, but Sunnabile 

 canes are slightly more so than Saretha. It was noted in examin.ing the canes 

 that a few of the thinii.er Saretha varieties had almost cylindrical canes — -a 

 rather rare pheu.omenon -and this probably accounts for the difference in the 

 averages. As to thickness in different parts, at nine mou.ths, the middle of the 

 cane is thickest, then the base, while the apex is thinnest. This thinness of 

 the apex is less pronounced i]\ the mature canes (1916 crop) an.d, indeed, 

 conditions are reversed and the apex is thicker than the base in both series. 

 This is especially noteworthy in the Sunnabile series, where the apex is thickest 

 in ripe canes and the base the thinnest. Thickening of the apex in Indian 

 canes is by no meau.s aii unusual character, and it is probably the Saretha 

 series rather than the Sun.n.abile which is exceptional in. this respect. 



(3) The variaiiott in Ihe length of joint in, different parts of the eane can 

 best be studied in the accompan.ying curve, in Chart II, which represent a 

 summation, of the curves of all the varieties in each group. As before, there 

 are two sets of measurements, those of the 1916 and 1917 crops, and in each 

 case 20 canes were taken in each variety and the length of successive joints 

 measured from ground level to the apex, until a joint only one-tenth of an 

 inch was reached. It will be obvious that the preparation of these curves has 

 entailed a very considerable amount of labour. To measure the joints in a 

 variety whose 20 canes average 32 joints each, means 640 individual measure- 

 ments, and tlie sununation series of 7 such varieties means 4,480 measure- 

 ments. The Sunnabile curve of 1916 is based on 8.200 measHremen.ts, while 

 that of the 1917 series is the result of a yet greater number of measurementti. 



