M. D. GhYjs'NK AND V. G. Jackson 



op, 



33 



variation and the differences in nianurial treatment, this high value 

 suggests that for many purposes determination of the specific gravity 

 may be a suflSciently accurate index of the dry matter content. 



For the calculation of the correlation coefficients we are indebted to 

 Mr E. M. Crowther. 



Method of Sampling. 



The unequal distribution of dry matter and nitrogen makes it difficult 

 to obtain a sample typical of the whole tuber. Coring, whether transverse, 

 longitudinal or diagonal, residts in the inclusion of too much of the central 

 wet part of the tuber and so gives too low a figure for the dry matter 

 content. The following table will give some idea of the error involved. 

 In six tubers treated individually dry matter was estimated in one quarter 

 and in a longitudinal core taken through half the tuber (Fig. 11). 



Fig. 11. 



Table XIII. 



Percentage of dry matter 



Average 



of 6 

 tubers 



21-43 

 15-81 



Part of tubor Tuber 12 3 4 5 (i 



One quarter 24-07 19-33 21-39 lS-43 21-34 24-04 



Longitudinal core 15-59 18-05 15-07 14-03 12-13 18-79 



In six other tubers a diagonal and a transverse core were taken 

 instead of a longitudinal one (Fig. \2). 



