120 TILLERING IN INDIAN SUGARCANES 



joints had not as yet reached one inch in length and were still in the process of 

 thickening, was exchided. The selection of this one inch length is pnrcly 

 empirical, bnt it answers the purpose well enough. The joints in this region 

 are very numerous and, towards the base, present the appearance of a series of 

 superposed discs, with difficulty separated from one another. We may 

 consider this portion of the cane as that engaged in attaining its full thickness 

 and in giving off branches, as contrasted with the following elongating portion, 

 and it presents certain characters which may be now briefly considered. As, 

 it is the formative region, in which the whole system of branches must arise, 

 it is not surprising that it is longer in the main stem than in its branches. It 

 has been measured in all the canes in each plant examined, and the table shows 

 the general averages of each group. We see there, that, while in the as of 

 all the groups its average length is 3-7", it is 2'6" in the 6s. W^e should expect a 

 similar difference between the 6s and the cs and so on, but this is not the case. 

 In fact, in most groups, there is a distinct increase in the length of the basal 

 part in succeeding branches after the 6s are passed. This fact has been already 

 explained by the presence in these later branches of curvatures, whereby they 

 may be placed in a better position for developing freely. And it has been 

 noted that, where there is such a curve, the joints remain short until the cane 

 has straightened out and is in a position to grow upwards in a vertical line. 

 In the Thick canes, in the Mungo group and in Saccharum, spontmieuw, there 

 is a continuous reduction in the length of the basal part throughout the series 

 of branches. This is readily explained in the first and last cases, in that curva- 

 ture is generally less evident in them, in the Thick canes because of their 

 comparative fewness, and in the wild Saccharum because of the thinness of 

 its branches, their general irregularity and the constant presence of runners. 

 In Saccharum spontaneum the shoots are placed in a position for free develop- 

 ment rather by runners than by curves, and the result is that the whole complex 

 of branches in the clump is loosely knit together instead of being closely compacted 

 as in the cultivated canes. In Mungo the case is different. There are a 

 great number of canes in the clump, but these arise from a large number of 

 separate buds, as many as 11 having been noted in one case. It is possible 

 that this even distribution of the individual plants along the length of the 

 set may serve the purpose of placing them, and thus do away with the necessity 

 of much curving. But curving is present, as also are runners, and the regular 

 decrease in the lengths of the basal portions of branches of successive orders 

 comes somewhat as a surprise. We have to wait until the da, before we get 

 the expected increase in length. With these two tendencies in opposite direc- 

 tions, we get, in the averages of all the groups, a uniformity in the lengths in 



