118 TILLERING IN INDIAN SUGARCANES 



most probable, as this plant has been grown for many generations in the south 

 of India, and indeed in the immediate neighbourhood of the farm, on richly 

 cultivated land as a fence round hetel {Piper Betle) gardens. The Pansahi 

 group seems to be also very hardy and little incommoded by the occasional 

 saltness of the land on the farm. Saretha varieties of the Katha section do 

 not grow well on the farm, and Saccharum spontanemn is not, at first, at home 

 in the cultivated land. The Mungo series is obviously at a disadvantage 

 because of the comparatively large number of plants per clump. The result 

 given in the table must therefore be taken as for one place only, with its many 

 peculiarities, in many cases not the best suited for free growth of the variety. 

 But it is improbable that any one place could be found where all the varieties 

 grown would be equally at home. With this word of warning, we can proceed 

 to analyse the averages in the table. The Thick canes, in the character of their 

 branching, show themselves furthest removed from the wild Saccharums. 

 Then come in order the Sunnabile and Nargori groups which approach the 

 Thick canes, then Saretha, Mungo and Pansahi, which are nearer to the wild 

 kinds. There is no reason to assume, only from the formulae above given, 

 that Saretha is nearest to Saccharum, spontanemn, although we have noted 

 many other similarities in Memoir III. The suggestion rather obtrudes itself 

 afresh that the Mungo and Pansahi groups of canes may have arisen indepen- 

 dently from some wild parent, and this agrees with the strongly marked charac- 

 ters of these types. The details in each group find no place here, and w^ould 

 fill up a large number of pages. A few notes on them are added at the end 

 of the Memoir. 



The above remarks refer to the formulae obtained from averaging the 

 canes formed at crop time. Shoots and burst buds and dead branches also 

 have their significance in the branching system, and we have included these 

 in a second series of columns. We may now turn to them to see if they show 

 anything of interest. It must be acknowledged, however, that the deaths have 

 not been very wisely marked down in the dissections. For them to be strictly 

 included in the branching system, only dead buds which had already burst 

 should have been counted. But while this was done at first, towards the end 

 of the work all dead buds were counted, on the assumption that they died in the 

 effort of growing out, which was probably often not the case. On considering 

 the enlarg3d formulae, we find that the general tendency is the same as that 

 in the canes at crop time. The Sunnabile and Nargori groups are now indis- 

 tinguishable from the Thick canes, the greater shooting in the latter being 

 possibly due to their encouragement in the process of ratooning or growing 

 for a second year from the same root stock. Some note may be taken of the 



