C, A. BARBER 113 



made less than the reality (c/. p. 127). The 8-10 months' clumps took at least 

 three months to dissect, but the same method was adopted, and it was fomid 

 that it was perfectly possible, at this stage, to decide what canes would 

 be matured at harvest time. This, however, applied only to the cultivated 

 canes. The wild Saccharums do not exhibit any special ripening, as for a 

 crop time, and the dissections made from six to nine months show great 

 individual variations in the rate of cane-formation. There was, in these forms 

 generally, an absence of large shoots which were not cane-forming at the 

 base, at whatever time the dissection took place, and in this respect they 

 were at variance with the rest. This fact shows that, in the cultivated canes, 

 the general plan of cane-formation for the season is laid down months before 

 the harvest, and the differences in the periods at which the dissections were 

 made were of little importance with regard to the end in view, which was to 

 obtain a scheme for each plant of the number and character of the canes formed, 

 with a general view of the shoots and buds and deaths occurring. The thick 

 canes were dissected late, on the assumption that cane-formation was verv 

 tardy ; but this has not turned out to be altogether the case, as will be seen 

 later. The matter is, however, of less importance, in that most of the thick 

 canes were dissected at crop time, and thus all immature canes were at once 

 rejected as not fit for cutting at harvest. 



The work was not without its special difficulties. The absence of well 

 grown representative specimens in some varieties in the dissection plots has 

 already been referred to. It is probable that, in these forms, not usually 

 growing well at the Cane-breeding Station, a more elaborate form of branching 

 may be more characteristic, as, for instance, in members of the Katha section 

 of the Saretha group, but this irregularity in the development of varieties 

 would probably occur at any one place where all the forms were being grown 

 together ; and in the present case it was merely considered sufficient to note 

 in the record that such and such variety was poorly gro-s\ii. In some cases 

 the branches were formed very near the base of the stem, and so intricate a 

 mass was revealed that it was almost impossible to get a connected picture 

 of the branching system. This was, for instance, the case with KagJize, which, 

 however, for other reasons, was rejected in the later dissections. In yet other 

 cases, the attachment of the canes at their base, whether to the original set 

 or to the later stems, was extremely thin and brittle. Such were bodily removed, 

 their places of insertion being marked by a series of duplicated pins with 

 numbers attached, and the general plant was then reconstructed after all the 

 sectors had been independently dissected. But a study of this firmness of 

 attachment showed at once that it varied greatly in the different groups, 



