C. A. BARBER 197 



■ (8) Width of leaf. This varies along the length of the leaf and, in order 

 to institute a just comparison, it has been found necessary to fix on tlie greatest 

 width, in the measurements. This was estimated by spreading out the leafy 

 shoot and selecting the widest leaf present. The figures therefore refer to the 

 extreme width of the leaf. In 1916 the average extreme width was TS" in 

 the Saretha. and TS" in the Sunnabile groups, while in 1917 the figures were 

 1*3" and 1 •?" respectively. As in the case of length of leaf , it is possible that, 

 here too, the widest leaves were still un.developed in the measurements of the 

 immature canes o£ 1917. The leaves in the Su]\nabi!e group are considerably 

 wider than those in the Saretha group. The Mesaiigan section a])pears to be 

 intermediate between the Katha section and the Sunnabile group, but it must 

 be remembered that it consists of bigger canes ; the figures in 1917 were Katha 

 section r2", Mesangan r5". 



(9) Leaf module. This expression signifies here, as elsewhere, length 

 divided by breadth. It is built up with the aid of figures contained in the 

 last two paragraphs ai\d is therefore average length divided by average extreme 

 width. As might be expected, from our knowledge that the organs are longer 

 and narrower in the Saretha series, the differences in the two groups are marked. 

 The 1916 figures are Saretha 35, Sunnabile 25, and. in 1917, Saretha 37 and 

 Sunnabile 26, resemblances sufficiently close to rule out all differences in vigour 

 of growth. In spite of the poorer gro\rth in 1917, the form of the leaf appears 

 to have remained i^ormal. The Mesangan section again occupies an interme- 

 diate position, the figures in 1917 being. Katha section 38. and Mesangan 33. 



(10) Leaf shape. These are the main measurements of the lamina. But 

 it was felt that the figures did not express obvious differences in shape. Most 

 of the Saretha leaves show a marked narrowing above the base and the widest 

 part of the Sunnabile leaf is. on the whole, lower down the length of the leaf 

 than in the Saretha grouj). A series of width measurements were therefore 

 made in the laboratory, in ten leaves of each variety, at different distances 

 from the base. These distances were empirically selected at 1", 6". 12" and at 

 the widest part, wherever that was, and its distance from the base noted. It 

 was surmised that these measurements would be sufficient for the purpose 

 and that, from them, we .should be able to form a correct idea of the form of 

 the leaf in each class. To complete the picture, it was necessarv to note the 

 full length of each leaf measured, but this was unfortunately not a])preciated 

 in the 1916 measurements, and the omission has been filled in by taking the 

 length of leaf obtained from the general curves of 20 canes measured in that 

 year. The following are summaries of these measurements, it being noted 



