l88 STUDIES IN INDIAN SUCAKCANE SEEDLINGS 



LEAF WIDTH AND TlLJ.EKJNll. 



Correlation 9. 



Tillering was ju{lt<ed by tlio total number of canes and shoots at crop 

 time. Both are entered in the table, allowance bein<i: made as before for 

 those removed during,' i)etty a]\alyses. There is a general tendency, marked 

 in the larger classes, towards a negative correlation between leaf widtii and 

 tillering. But there is a remarkable exception in the case of the Sorrl/Hi 

 series, where there is a, just as stroiigly marked, positive cori elation. This is 

 unexplained, but it is worth while remembering that this series is j)j'actically 

 the only one of indigenous North Indian canes, ai\d furthei studies may sliow a 

 ditt'erenc<^ l)etween them and the tropical o]\es in this respect. It must also 

 be i'emend)ered that it was ])ossible to divide up the t^arelha seedlings i)ito 

 habit classes (p. 1.'33) and, o]i examination, it will be seen that the class with 

 biggest seedlings had nuich the greatest weight and also greatest leaf width. 

 It is possible that these habit classes, heie so easily seen, dominate, and that 

 they are absent in the thicker canes. The general result is. howevc. what 

 wo might expect on contrasting the tilhning of tro|)ical and North Indian 

 canes, but the differences are much sniallo!' than in that case. In th(^ above 

 remarks I have igiuned the small Poorati class, for an examination of the 

 tillering figures in the series shows the greatest fluctuations and flie total 

 lutmbcr of scedlint^'s is very small. 



