C. A. BARBER 177 



absent in adjoining canes, but present in canes slightly removed from one 

 another, as in Nos. 8-10 and 8-11, the connecting lines are broken ones. 

 Lastly, a dotted line is introduced towards the right of the diagram, which 

 gives a general curve of the length of the twenty canes, as judged by the 

 number of joints in each. 



Now, it is interesting to note that the lines connecting related maxima 

 in two adjoining canes are often more or less parallel, with one another and 

 with the dotted line on the right. Parallelism can be traced, for instance, in 

 the connecting lines between Canes 3 and 4, i and 5, 5 and 6, 7 and 8, 8 and 9, 

 12 and 13, and few adjoining canes are without traces of them. By somewhat 

 arbitrarily joining up all the connecting lines throughout the series, we get a 

 set of maxima curves for the whole twenty canes. In composing them, atten- 

 tion is paid to the undoubted tendency of the connecting lines to run parallel 

 to the dotted line on the right. Although showing numerous irregularities, 

 the general curves of maxima thus obtained are readily seen to be more or 

 less parallel in their course with the curve indicating the length of the canes. 



If the periodic stimuli indicated by these curves synchronize, it is natural 



to suggest their origin in some external cause, acting on all the canes together. 



If, on the other hand, they do not occur at the same time, we are thrown 



back on a natural periodicity of growth in the plant, which, in this case, tends 



to form longer joints once in every five or six. But we have some knowledge 



of the relative time of origin of these twenty canes. We have learnt elsewhere 



that the canes of a clump are often easily distinguishable into two classes, 



early and late in origin, and this shows itself especially clearly in the Ponsahi 



group of canes. ^ Among a number of distinguishing characters, the first 



formed canes have a considerably larger number of joints than those 



emerging later from the ground. We are justified in assuming from this 



that the upper canes in the diagram, having the greater number of joints, 



are early canes, and those towards the bottom of the diagram are late. 



Arguing from these premises, we may suggest that the first four canes, being 



the earliest formed, have a very early maximum in length of joint. The next 



eight, arising somewhat later, have a slightly later first maximum. It is 



possible that the first maximum in these eight may synchronize with the 



second maximum of the first four, and this is independently suggested by the 



maxima curves. Similarly, the next four canes have a still later first maximum, 



and the connecting lines suggest that these synchronize with the second 



maxima of the second eight and the third of the first four. Lastly, the last 



1 Barber, C. A. Studies in Indian Sugarcanos, No. 2. Sugarcane seedlings, including some 

 correlations between morphological characters and sucrose in the juice. Alem. Dep, Agric. , 

 Ind., Bot. Ser.. Vol. VIII, No. 3, July, 1916, p. 159 



