C. A. BARBER 



75 



the soil are lacking. Kilian's results on the numbers of canes are summaiized 

 in the following table : — 



Numbers of canes per howw (If acres) with different manures. 



* The words "slight fall," "rise," etc., below an average figure, indicate any changes 

 within the bracketed tieatments. 



Kobus^ in an earlier jiaper (1905) describes the results of his experiments 

 on growing cane uninterruptedly on the same land for a succession of years, 

 with a various assortment of manures designed to take the place of rotation 

 and fallowing. In this series were N, N and P, N and K, N, K and P, and 

 all these with or without the previous addition of Ca. The series is a very 

 full one, as it deals with three different varieties of cane, J. 247, J. 33a and J . 36. 

 He states that the plots were much affected by the weather, there being a 

 severe drought in the earlier part of the season, but that they recovered much 

 better than he had expected. There were, however, many failures in germina- 

 tion, varying from 6-4 per cent, to 12-4 per cent, in the different varieties. Rats 

 invaded the plots and created great havoc, to an extent in some cases of 40 

 per cent. Lastly, J. 36 suffered from red rot, as this variety is more liable to 

 the disease than the others. Among other data, he obtained the number of 

 canes in each plot, and his general conclusions were that tillering is com- 

 paratively unaffected by manuring. By this we presume that he means rather 

 that the kind of artificial manvre a'pplied, whether nitrogen, potassium, 

 phosphorus or calcium in their various combinations, has little effect on the 

 number of canes produced per acre. This does not seem to be ciuite the same 



1 Kohus, J. I). Cultuur van suikcrrict zonder tusschcn-gcM'asscn. Archirf y. d. Java, 

 Huiktnnd.y Vol. XIII, 19U5, p. 485. 



