C. A. BARBER 89 



increase in the differences in the numbers of shoots produced per plant 

 as the period of harvest approaches, which is not to be wondered at, as 

 the effect of the spacing should be cumulative thioughout the growth of 

 the plant. 



The same author conducted spacing experiments on a very large scale, 

 the plots extending over 100 bouws (175 acres) with sets planted roughly 

 as 2 to 3 for the same space. This again is a smaller difference in space allowance 

 than Stubbs's, but the results are obvious enough. The numbers of canes per 

 bouw are practically equal, showing that the effect of the spacing was that 

 each clump, on the average, produced half as many canes again in the wider 

 planting. 



Wider spacing thus has a marked influence on the maxinuun number 

 of shoots developed per plant ; this effect is cumulative, during the period 

 of growth, and is therefore intensified at the time of harvest. 



(c) Effect of sixicing on the thickness or weight of the individual cane. 

 Stubbs gives the average weight of cane when the sets were planted 18", 12" 

 and 6" apart, in lb. as 2-60, 249 and 2-17. Kilian's results are less conclusive, 

 but the distances apart in the 3|', 4' and 5' rows were very much less. 

 The relative weights in the two tables were as 3"2 to 3'3 to 3"35 and 

 3"8 to 3'75 to 3"85. There is thus practically no difference in the weights 

 of the canes. Muller von Czernicki dealt rather carefully vn\\\ the thickness 

 of the cane, and he deduced the weights on the assumption that the canes 

 were of equal height (which he states from observation is not perfectly 

 correct). He measured the canes at 5-6 months with calipers, in the rows 

 with 8 and 18 sets in them. The result that he obtained from a large 

 number of plots was that the canes in the 8 sets plots were 14 per cent, 

 heavier than those of the 18 sets plots. Other observers, notably Kobus 

 and Van der Stok, emphasize the fact that wider spacing increases the 

 thickness of the individual canes, and it may be considered therefore as 

 incontestible. 



{d) Effect of spacing on total iveight of canes at harvest. A wider spacing 

 therefore produces more canes per plant, and these are thicker and heavier. 

 But there are fewer of these plants to the acre. Stubbs gives figures for the 

 total weight of canes reaped, with his spacing of 18", 12" and 6" in the row, 

 as 37"24, 41*6 and 42"55 tons per acre, a distinct though small advantage for 

 the closer planting. Kilian's figures agree, taking the smaller differences into 

 account in his spacing experiments. The total weights of canes in the 3|', 4' 

 and 5' rows were, in pikuls per bouw, 2070, 2056 and 1978 respectively. 



