MANURES 95 



Thus, when potash is omitted, the result is as bad as when 

 no manure at all is applied : the omission of nitrate has no effect, 

 the results obtained in that case being as good as when the 

 complete manure is applied : and the omission of phosphates 

 is actually beneficial, to the extent of about 17 per cent. 



These results are entirely in accordance with what might be 

 anticipated from the character of the soil as indicated by 

 analysis (p. 85). This showed that there was a considerable 

 deficiency of readily available potash namely, 0*006 per cent., 

 against 0*025 P er cent, at Ridgmont but that as regards phos- 

 phates, the Millbrook soil was rich, and actually richer than the 

 Ridgmont soil, containing 0*024 per cent, of phosphoric acid, 

 against 0*014 per cent, at- Ridgmont. Since it is, therefore, 

 available potash which this soil requires, manures will do no 

 good unless they contain that element, and, when it is omitted, 

 the results are, as has been seen, just as bad as when no manure 

 at all is given. But phosphates are already present in sufficient 

 quantity in the soil, and the addition of a further amount of 

 them in the manure is not required ; indeed, we find that if we 

 supply the potash without the phosphates (" phosphates omitted ") 

 we get even better results than if we apply both together (" com- 

 plete manure"), for, in the absence of the phosphates, the 

 relative proportion of potash in the total food available for the 

 plant will be greater. 



That nitrate should have had little or no effect on the results 

 at Millbrook is not surprising, in view of the fact 'that the soil 

 there is one of the coarsest of sands, from which nitrate of soda 

 would very readily be washed out ; added to which, it already 

 contained a fair supply of nitrogen, namely, 0*1 per cent, in the 

 top nine inches, as compared with 0*136 per cent, at Ridgmont. 



RESULTS WITH GOOSEBERRIES (Reports, IV, 35 ; XVI, 29) 



There were two main series of experiments wiin gooseberries, 

 the plots consisting of about 200 representatives of all the 

 principal varieties. In one series, A, the records extended from 

 1897 to 1911, inclusive, in the other from 1912 onwards. The 

 main feature of the results in both cases was the extraordinarily 

 great effect of dung; this, however, did not, in the case of the 

 first series, begin to be very apparent on the crops till after the 

 fourth season, the effect till then being mainly one on the growth. 

 To obtain a fair estimate of the results from the fruit crops, it 

 is well, therefore, to omit the records for these first years, and 



