C. A. BARBER l63 



grown at Saboiir in 1917 (Plate I). Besides these measurements in length of 

 organs, the numbers of joints are counted and the total length of formed 

 cane is measured, and from these two figures is obtained the rate of cane 

 formation per mensem while in the ground. The numbers of dead leaves are 

 counted and the length of the cane covered by these (these two figures being 

 taken as indicating the rate of ripening of the cane), the length of the shoot 

 (unripened part of the plant) is recorded and the total length of the whole 

 plant when laid out on the ground, the maximum width of the leaf and the 

 average thickness of the stem. The latter measurement deserves special 

 mention, in that canes vary in thickness in different parts and are generally oval 

 in section. Each cane, when stripped, is measured at the middle, at the base, 

 and at the highest matured joint. And these measurements have to be done 

 in two planes, at right angles to one another, because of the ovalness of the 

 cane, one being made in the plane of leaf attachment and one at right angles 

 to it. The latter is almost invariably narrower than the former. These six 

 measurements are averaged for each of the twenty canes, and the resultant 

 of these is taken as the average thickness of the canes in the plot. Incidentally, 

 ovalness of the cane is recorded, as the cane varieties differ in this respect, 

 as well as the tendency to thicken upwards or downwards which characterizes 

 different kinds of canes. All of these measurements are made on the same 

 twenty canes, and the latter are so arranged in tables that all the measurements 

 of each individual cane and joint may at once be seen, and any future work 

 on correlation between the variations in size in different organs may be 

 studied. 



This material forms the basis of our work, and a comparison of these 

 measurements determines the general lines on which our conclusions are 

 framed. It becomes possible to compare the gro^^i}h of the same cane in 

 different places, the development of individual cane characters in different 

 surroundings, and the manner in which different sets of conditions impress 

 themselves on cane gro^vth in general— whether favourably or the 

 reverse. 



The series of measurements obtained for each set of twenty canes, taken 

 at one time and place, form together a " unit " observation. There are now 

 some hundi'eds of these units collected, and many of them were utilized in 

 the Memoir referred to above, for the purpose of distinguishing the gro\\i:h 

 characters of the Saretha and Sunnabile groups. But, in the present paper, 

 only those dealing with, the seven cane varieties distributed to Taliparamba 



