MANURES 



91 



and the average size of the fruits (p. 19). 

 measurements gave the following results 



A summary of the 



There are here three features which show an advantage in favour 

 of the highly manured plots, and three which show the opposite : 

 there are only two the weights and values of the crops which 

 show any regularity in the variations of the results of the three 

 groups with the amounts of manure received by them. If the 

 annual growth-measurements are omitted (which apply to the 

 young trees of Gascoyne only), the mean values (last line) 

 show a slight decrease with a reduction below normal of the 

 manure applied, but no appreciable increase (only 0*2 per cent.) 

 with an increase of that manure above the normal. The variation 

 observed, however, must be regarded as insignificant, for on 

 determining the probable errors of the results in the usual way, 

 those of the two groups under extreme manurial treatments 

 amount together to 2*4, whereas the differences between the 

 means for these groups is only rg; that is, the means are 

 identical within the limits of experimental error. The conclusion 

 to be drawn must be that differences in the amount of manure 

 have produced no measurable differences at all. 



A similar conclusion was drawn from results obtained in two 

 mixed plantations of one-quarter of an acre, each containing bush 

 and standard apple trees, with a few pears and plums; these 

 had, since 1905-6 eleven years after they had been planted 

 been subdivided into sections receiving different manurial treat- 

 ment. The average gross values of the crops for the following 

 ten years showed in one case (A) x a considerable balance against 

 manuring either with dung or artificials, whereas in the other, 

 though there is a balance in favour of the more heavily dunged 



1 The values given in the Sixteenth Report, p. 23, are erroneous, owing 

 to a miscalculation. 



