THE PRINCIPLES OF MANURING 191 



in the same material. The discs are perforated with 

 holes 'to allow the plants to grow and through which 

 water may be added. The pots are weighed every two 

 or three days in order to measure the loss of water by 

 transpiration. At the end of three or four weeks the 

 plants are cut and weighed and the amount of dry 

 matter in them is estimated. Both sets of data are con- 

 sidered in estimating the effects of the fertilising in- 

 gredients, and the results correspond closely. There are 

 many possible sources of error in the conduct of the 

 operation and great care is required to secure com- 

 parable results. In any case the method cannot be 

 considered as reliable as field trials. Its chief advan- 

 tage lies in the comparatively short time in which it can 

 be carried out. 



The illustration 1 ,] (Fig. 27) shows the results obtained 

 in two tests applied to the same soil ; the manures were 

 the same, but different kinds of plants were used that in 

 the upper figure being rape, and that in the lower, clover. 



Chemical Analysis. The manurial requirements of the 

 soil may also be determined by chemical analysis. For 

 this purpose the investigation may be confined to the esti- 

 mation of the total nitrogen and available phosphoric acid 

 and potash. It is quicker than field trials and the results 

 are not obscured by the peculiarities of a growing crop. 

 It may be easily extended so as to tal^e into account other 

 considerations, and it has sometimes succeeded when the 

 eight-plot test has failed. It is more expensive than the 

 11 practical " test and is generally considered not so reli- 

 able. So long as a suspicion of this kind remains, the 

 conclusions arrived at by chemical analysis must be subject 

 to confirmation by experiments in the field. The two 

 methods may therefore be regarded as supplementary. 



1 Bui. 2, New Zealand Dept. of Agriculture. 



