C. A. BARBER 



65 



an end row was taken and this is noted in the table. It was presumed that, 

 because of greater space, there might be more canes in an end than in the 

 central row, but this does not seem to be so to any great extent, and the position 

 of the row may accordingly be neglected, as of little influence on the general 

 results. 



The number of clumps examined in these observations is therefore 

 extremely limited, especially as the end clumps of each row, that is, those next 

 to the irrigation channel and the drain at the other end, were excluded. The 

 table is less to be relied on than the preceding one, but the general results agree 

 well enough with those obtained on the farm in other years. The average thick- 

 ness of the canes in the plot has been added, and this was computed from the 

 average of 100 canes in each case. In this table, the varieties in each group are 

 arranged in the order of greatest tillering. Comparing the average rates of 

 tillering, we see that Mungo and Nargori groups head the list with 15" 1 canes per 

 clump ; these are closely followed by the brown section of Saretha, then, in 

 succession, the series of unclassed canes, Sunnabile group, Pansahi group and, 

 last of all, the green section of the Saretha group, with 9'5 canes per clump. 

 As regards thickness, Mungo heads the list and Nargori shows the thinnest 

 canes. But both of these have the greatest tillering power, a fact that has 

 already been commented on. The following summary table shows that, with 

 the exception of Mungo and the gresn section of Saretha, the average thickness 

 of the canes in a group varies more or less inversely with the rate of tillering. 



