C. A. BARBER 



139 



In conclusion, the depressed habit referred to above is not an uncommon 

 occurrence. Many of the Gramminese show two markedly different growth 

 stages, the first consisting of much branching of low shoots and the second of 

 the erection of flowering spikes. But this habit of growth is not altogether 

 disadvantageous, for it does not diminish the number of ears of grain borne 

 aloft at harvest time. In the sugarcane the inflorescences are also erect, but 

 that is of little importance, and the twisted nature of the stem is n matter of 

 real concern. 



Tillering. 



Tillering varies a great deal among seedling canes, and the cause of this 

 is not understood. Many of them, especially among the thicker canes, at 

 first develop a single cane-like shoot which becomes quite large before any 

 others are formed. Others again send out a number of smaller equal shoots 

 which continue to grow for some time before one of them becomes a leadmg 

 shoot and outstrips the rest (Pi. XXIV. See also Pi. XVI). The two 

 extremes are usually only met with in the Indian canes, where we get, on the 

 one hand, a single cane-like shoot and, on the other, a dense mass of small leafy 

 ones with thin leaves, for all the world like a tuft of grass (c/. PL XIX). The 

 latter form is not generally found in the seedlings of thick canes where, however, 

 it is not infrequent for a seedling to develop a number of tallish, graceful 

 branches so as to resemble a full grown plant of tenai {Setafia italica). No 

 connection has at present been established between the number of shoots 

 early developed and the mature habit of the seedling but, as a rule, seedlings 

 of Indian canes have, like their parents, far more branches than the thick 

 tropical ones. This is readily seen in the accompanying table. 



Tillering in seedlings of vafiow? parentage. 



