8^ TILLERING IN INDIAN SUGARCANES 



reaped, but here also we are in the dark as to whether this is accompanied bv 

 a diminution of the relative numbers of cs and r/s, which might give the 6s a 

 predominance over them. This side of the question is of some importance, 

 for we shall see later that, while the as and 6s are often very similar, the cs and 

 ch differ radically from them, and the question is thus not only influenced by 

 the relative number of as in the plots, but also whether the as and 6s on the 

 one hand or the cs and (h on the other are in relative excess. Until we have 

 more definite information on these points, we should prefer to rely on the actual 

 analyses of early and late canes as given by Stubbs and carried out for some 

 years on the Cane-breeding Station at Coimbatore ; but these latter will be 

 referred to later. 



Muller von Czernicki states that he has often noted tbe differences in 

 thickness of canes sparsely and closely planted, especially in the 1909 experi- 

 ments, and he decided to test this more carefully. He therefore measured 50 

 canes from each plot in the following manner, making altogether 1 ,000 measure- 

 ments. He used a pair of calipers which he moved round the stem until it 

 encountered the greatest resistance, and took the measurements at about 

 one metre from the ground at the middle of an internode. His results are 

 given in a series of tables, in which the canes are arranged according to their 

 thickness in each plot, with differences in millimetres. From these measure- 

 ments in thickness he deduced the weight of the canes. By using a formula he 

 calculated the difference in average weight of canes in the rows with Sand 18 

 sets, the extremes of the series. This difference varied from 10*5 per cent, in 

 imported Cheribon sets, to 17"6 per cent, in local J. 247. The average of 

 these differences in the four kinds of seed used was 14 per cent., which 

 means that 86 canes in the thinly sown rows would equal in w^eight 100 in the 

 closer planted rows. In these deductions he assumes that the plants in the 

 different rows were of equal height, but he himself observes that this was by 

 no means the case ; he therefore concludes that for accurate determinations 

 of the weights of individual canes direct weighings will alone suffice. 



Striiben, in his paper on Tillering (1911), already mentioned, collates 

 numerous countings of canes made by different workers, under the most varying 

 conditions of climate, soil and treatment, and concludes that, within narrow 

 limits, each variety shows the same cane-producing capacity, limits narrow 

 enough not to be of appreciable influence from the crop point of view. He 

 further gives the results of a series of experiments conducted by himself on the 

 lines laid down by Muller von Czernicki. He experimented with J. 247 and 

 placed 6, 8, 10, 12, 14. 10 and 20 sets in separate rows of the same length, 



