488 On the Groicth of Barley hy different Manures, Sfc. 



increase, obtained by the use of nitrogen equal 100 lbs. or more 

 of ammonia per acre per annum, as by the 50 lbs. or its equiva- 

 lent of nitrogen. Bearing in mind this general result of an ex- 

 cessive amount of nitrogen supplied, the figures are still deserv- 

 ing some examination. 



In this Series 4, the nitrogen was provided in the three sepa- 

 rate forms of nitrate of soda, an equal mixture of sidphate and 

 muriate of ammonia, and rape-cake. In Series 3, the quantities 

 of ammoniacal salts and nitrate of soda em.ployed per acre w^ere 

 100 lbs. each, sulphate and muriate of ammonia, and 275 lbs. of 

 the nitiate ; the 200 lbs. of mixed ammoniacal salts and the 

 275 lbs. of nitrate of soda having been taken as containing re- 

 spectively in round numbers 41 lbs. of nitrogen, = 50 lbs. of 

 ammonia. 



In Series 4, double the amounts, maintaining therefore the 

 same proportions to one another, of the ammoniacal salts and 

 nitrate, were employed. These quantities were adopted on the 

 assumption that the equal mixture of the ammoniacal salts, in the 

 condition in which they are sold in commerce for manure (though 

 we always use the best at our command for experimental pur- 

 poses), will average about 8 per cent, of water and impurity, and the 

 nitrate about 8f per cent.* It is probable, however, that the 

 nitrate will frequently come up nearer to the chemical standard of 

 composition than the ammoniacal salts, and if so, our supposed 

 equivalent amount of nitrate will contain a correspondingly larger 

 amount of nitrogen than the ammoniacal salts ; and this may 

 partly, though not entirely, account for the greater effects ob- 

 tained by the nitrate employed. The amount of rape-cake 

 employed in Series 4 was 2000 lbs. per acre per annum ; and it 

 is assumed in the Tables, and also sometimes when speaking of 

 the results, that the amount of nitrogen thus supplied was about 

 the same as that in the larger amounts of ammoniacal salts and 

 nitrate used by its side. This would, however, suppose a lower 

 per cent, of nitrogen in the rape-cake than the average of our 

 own determinations, or the records of those of others, would indi- 

 cate. It is, indeed, not improbable that the 2000 lbs. of rape- 

 cake would frequently contain nitrogen equal to from 110 to even 

 120 lbs. of ammonia. It is probable, on the other hand, that 

 ihe nitrogen of tlie rape-cake will be given up from its existing 

 .d:;ombinations, in less proportion each year tlian that in either of 

 the other two forms. Although, therefore, all the experiments in- 

 cluded in Series 4 are so brousfht tosrether for the convenience of 



* These deductions are probably in both cases rather excessive. It frequently 

 happens, indeed, that sulphate of ammonia in the bulk, will average less water of 

 crystallization than is theoretically accorded to it. 



