C. A. BARBER 43 



will throw some light on the stage of development of each group from its 

 supposed wild ancestor. 



The differences in form and size betw^een the branches of different orders 

 in the same plant have been carefully studied. Each cane has been measured 

 aa to the length of the basal, branching portion before it has assumed its full 

 thickness, the thickness at two feet from the base, the average length of the 

 joints in these two feet, the presence of curvature and runners, and so on. 

 In all of these characters we find, as might have been expected, that there is a 

 tardier development in the first shoot, and that this development increases in 

 rapidity as the branches of the higher orders are reached. The general trend 

 is for the branches of higher orders to be thicker, to have longer joints, and to 

 show greater curvature. The main shoot has a longer basal, preparatory 

 portion than its branches, but, when we pass to the other orders, the presence of 

 basal curvatures, needed to place them in a position for upright growi;h, again 

 increases the region of short joints at the base, for it is the general rule that a 

 branch does not assume its full form until it is in a position to grow" onward 

 unimpeded. These details are all arranged in tabular form for the variety or 

 group in the body of the Memoir {cf. Part III, Sections 4-6). In Memoir III, 

 an example was given showing some of these characters, w^hich was illustra- 

 ted by plates. Here I add a more striking instance, namely, that of an ordinary 

 plant of Saccharuni arundinaceum (Pis. I and II), a species marked by its 

 upright habit and symmetrical development, just as Saccharuiti spontaneum 

 is by its intricate spreading growth. The diagram and detailed table of 

 measurements are added, to give some idea of the character of the work done 

 in each dissection undertaken. For further results, the reader is referred to 

 the body of the work, as it is impossible adequately to summarize them 

 without undue repetition. 



The characters of the branches of different orders are seen to be so definite 

 that, when a field is cut, w^e can without difficulty separate the canes at the 

 mill into early and late. There is a good deal of similarity sometimes between 

 the OS and 6s, especially wiien the latter become facultative as, but the change 

 from 6s to cs and ds, is sufficiently striking to render their distinction generally 

 very easy. This opens up a new- line of work, in that it becomes possible 

 to analyse these branches separately and to settle the question of their relative 

 richness of juice and other qualities at the mill. Most of the work has, it is 

 true, been conducted with Indian canes, and in one particular locality, but it 

 seems unlikely that the thick canes will not fall into line, considering the general 

 similarity of their branching system to that of Indian ones. 



