C. A. BARBER 



121 



6s, cs, and ds, which is rarely met with in the individual plants. As in other 

 cases, we have to turn to Saccharum arundinaceum for the complete regularity 

 of the series. Here the lengths in a average 2'9", and there is a sharp drop to 

 the 6s with 2"1", followed by increases to 2'4", 2"9", 3*2", in the succeeding 

 branches of higher orders. A glance at the uprooted clump in this species 

 will show the increasing curvature which is a dominant factor in this character 

 {cf. PI. I). 



Average length of hasal part, with joints under \" in length, i 



in inches. 



(5) Average length of joints in branches of different orders. 



The main shoot arising from a bud on a set is different from its 

 branches in several respects. It has been shown that it takes longer to 

 develop, in that the plant is small and weak at first ; it has a longer basal, 

 branching portion ; it is also markedly shorter jointed than its successors. 

 This will be seen from the Table, where the lengths of joints in the os, 

 6s, cs and branches of higher orders have been averaged for the different 

 groups. It may be noted that the numbers of branches dealt with are not 

 the same as in the formulae of matured canes, for even immature canes have 

 been included, if it was possible to measure the lengths of the joints in the 

 first 20". It is well known that the lengthening of these joints, once they 

 are formed, is extremely small and may be neglected.^ 



The averages have been obtained in three ways, the as, 6s, etc., of each 

 plant have been averaged separately, the averages for the varieties have been 

 obtained from these, and these latter have been again averaged for the groups 



^ Kuijp3r, J. De groei van blad-schijf, bladschcede on stengel van hot suikcrriet, Med. 



Java Snikeri rid.. V, 8, 1915. 



