C. A. BARBER 49 



from lack of room, appear all together as an irregular circle round the main 

 shoot, but careful dissection shows that they all arise from different axils on 

 alternate sides of the plant (PL VI). The growth of successive buds, how- 

 ever, varies a good deal, and their size at this stage bears no sort of relation 

 to the time at which they were formed at the apex of the stem. Some buds 

 remain quite small during the life of the plant, whereas others grow so rapidlv 

 that they soon overtake or even exceed the main shoot in length. 



The branches pass through exactly the same stages as the parent stem, 

 only differing from it in that they have a better start and take less time to 

 develop into leafy shoots. They are thin at the place of origin, bear closelv 

 packed leaves on the short congested joints, have a bud in the axil of each leaf, 

 and, as the leaves increase in length and expand their blades, the stems increase 

 in thickness, the successive joints become longer and the shoots as a whole 

 emerge from the ground. As in the main shoot, the leaves at first grow much 

 faster than the stem and, for a long time, the actual growing points of the stems 

 remain below the ground, the height of the plant being judged by the length 

 of the expanded leaves. This is readily explained by the fact that the growth 

 of each shoot is largely dependent on the feeding power of its own leaves and. 

 until these are fairly large, no real progress can be made, hence their earlv 

 protrusion and proportionately rapid early growth. The relative size of the 

 main shoot and its branches, and the number of the latter vary much in the 

 same batch of seedlings, all stages being observable between one strong cane 

 with or without a few small shoots at its base, and a bunch of shoots resemblin"- 

 a tuft of grass, in which it is difficult to distinguish betw^een the main stem 

 and its branches (c/. Pis. XVI and XXIV of Memoir II, for illustrations 

 of this). The reason for this is not clear, for seedlings thus differing in their 

 early stages are often not distinguishable in their degree of branching later on. 



At a somewhat later stage, the lateral shoots, each as fully provided with 

 buds as the parent stem, may also branch, giving rise to branches of the second 

 degree, and this process may continue to several further degrees, this depending 

 to a large extent on the parentage of the seedlings. Such shooting of the 

 buds on lateral branches is not, however, usual until the plant has reached a 

 further stage of development, unless, indeed, one of the branches receives 

 an accidental injury low down, when its place is often taken by one of its 

 uppermost buds. The chief points to bear in mind, concerning the branchin<' 

 of cane seedlings, are that every joint has its leaf and, protected by it, a bud, 

 that both joint and bud have the power of forming independent roots if the 

 necessity should arise, and that any of these buds may remain quiescent or 



