166 PROCEEDINGS OF THE CHEMICAL DEPARTMENT. 



cereals, is particularly worthy of notice, and raises a question of 

 much practical importance. For some years back Peruvian 

 guano has been largely used as a top dressing for cereals, and 

 particularly for oats, in spring, and is by many peasons believed 

 to be the most economical way of using that manure ; but if, as 

 these experiments lead us to conclude, it acts by virtue of its 

 nitrogenous constituents alone, its mineral matters are so much 

 valuable matter which is either lost or fails to make any return 

 until a later period. Now, in buying a ton of Peruvian guano 

 at £13, the farmer pays £10 10s. for the ammonia (reckoning all 

 its nitrogen in that form) which it contains, and £2 10s. for its 

 mineral matters ; and if the latter is to produce no immediate 

 effect, it would in all probability be better for him to expend his 

 money on sulphate of ammonia, because he can obtain for £10 

 10s. a quantity of that salt sufficient to yield the same quantity 

 of ammonia as a ton of guano. 



In addition to the conclusions to which they lead regarding 

 the questions originally proposed for solution, these experiments 

 offer an interesting illustration of the advantages attending the 

 use of small plots, on which I insisted very strongly in my 

 address on "Experimental Agriculture" at the Stirling Show. 

 "It would, in fact, have been scarcely possible to make such 

 experiments on the ordinary quantities of \ or even \ of an acre. 

 The preparation of uric acid, free from ammonia, even for the 

 small plots to which it was applied, was a sufficiently laborious 

 and troublesome process. The cost of the materials required to 

 produce a ton of uric acid amounted to about £80, to which the 

 cost of manufacture would have to be added, but this process 

 was conducted in the laboratory,and cannot be exactly estimated ; 

 I have little doubt, however, that, had a manufacturer been 

 employed to make it, the price would not have been much under 

 £100 a-ton. The guano ash cost about £10 a-ton, so that it is 

 sufficiently obvious that these experiments, if conducted on a 

 larger scale, would have been so costly as to have deterred any 

 one from attempting them. 



They show also the utility of interspersing numerous nothing 

 plots among those to which the manures are applied, and the 

 facility which this affords for detecting variation on the quality 

 of the soil. This is remarkably seen in the experiments in oats, 

 where a difference so trifling as to be scarcely deserving of 

 notice, is made distinctly manifest. In the hay experiments, 

 where the variation is larger, the multiplication of nothing plots 

 Lrives the most conclusive evidence that the mean of the dupli- 

 cate plots is the true average, and thus enables us to draw our 

 conclusions with much greater confidence than could otherwise 

 be possible. 



In many other respects, also, the advantages of small experi- 



