122 TILLERING IN INDIAN SUGARCANES 



(column 1). Secondly, all the an, 6s, etc., of each variety have been averaged, 

 and these variety averages have given those of the group (column 2). Thirdly, 

 all the. rts, &s, etc., of each group have been taken together (column 3). It is 

 remarkable how little difference there is in the three resulting sets of figures. 

 There is a steady rise in the average length of joints in every single group, 

 and sometimes a sudden increase between the 6s and cs. The reason for the 

 selection of the two lowest feet of the cane for these measurements has been 

 already given in detail (p. 106). 



In the Table, where there are less than four measurements of a branch 

 of one order in a group, the resulting average is placed in brackets, as insuffi- 

 ciently reliable for the group average. Such figures are however included in 

 the total summation at the foot of the Table. If less than 20" can be measured, 

 the cane is rejected as immature. The joints of runners are omitted and the 

 lengths of joints are not measured until the short-jointed portion following 

 the TUnner is passed. If moth-borer or other injury is noted as reducing the 

 length of the joints above it, the cane is rejected ; this is judged by comparison 

 wnth other branches of the same order in the same plant. 



