900 



0.4 inch less thau that from a soil 20 inches deep; the amount of drain- 

 age was gioatest in December and least in July; the proportion of 

 nitrates in the drainage Avas generally greatest in September and least 

 in March, this proportion being much more uniform throughout the year 

 in the 00-mch thau in the 20-iuch soil; the total quantity of nitmgen as 

 nitrates removed in the drainage water was greatest in November and 

 least in April. 



lu the second case it was observed that when the spring is a forward one, as in 

 1884, the nitrates may dieapjiear from the drainage of the nnmanured wheat plat in 

 March; in other years the disappearance has occurred in April or May. In June it 

 is rare to find nitrates in this drainage water. * * * In September, the crop 

 being now removed, nitrates are always found in the drainage water. In a wet 

 season the maximum amount of nitrates will occur in Oitober. The proportion of 

 nitrates will be maintained with little diminution during the winter months and 

 begin to fall again in March. » * * Those facts not only illustrate the etfect of 

 a growing crop in reuioving nitrates from the soil, but they also point out the gn-at 

 defect of cereal crops, consiilered as conservers of soil nitrogen. The active growth 

 of cereal crops ceases generally in .Inly and they are entirely removed from the land 

 in August or .September; they are thus unable to att'ord protection against loss of 

 nitrates during the aiitiniin months, the .sea.s<ui at which the greatest losses occur. 

 From this point of view maize is a more economical crop than either wheat, oats, or 

 barley, its growing period extending through the whole of the summer. Nor must 

 we forget that the need which a crop has for nitrogenous manure is in inverse pro- 

 portion to its own power of assimilating nitrogen. After a wet winter cereal crops 

 begin to grow in a soil iuipovcrished of its nitrate**, and tl«e growth of most cereal 

 crops is over before the summer production of nitrates is half accomjilished. Cereal 

 crops are thus especially bi-nefited by nitrogenous numures, and particularly by the 

 application of nitrates, wlule. for tlie reason already given, maize is more ind<'pend- 

 ent of such manuring tli.in wheat or barley. The beneficial influence of a dry 

 winter upon the croi>s of the ensuing year is now generally recognized. 



Experiments inv a nnmltcr of years with h'gnminous crops, the 

 results of which arc tabulat«'(l. indicate "that the average proportion of 

 nitrates in the drainage water i)a.ssing through a soil growing leg:umi- 

 nous crops will generally be grealer than in cas«'swliere cereal eroj)s are 

 grown. * • • The drainage water being richer, the proportion of 

 nitrates in the subsoil out of reach of the changes produced by crop 

 and season, will also be greater.'' 



The (lesirability of inaetieing rotations whieh keep the land under 

 crop continuously in order to conserve the nitrates, is urged. 



The amonnl of nitrates in the drainage water from plats eroppe(l with 

 wheatand variously manured, and in the soil of plats ^ ariously manured 

 and cropped with wheat and barley, are tabulated and discussed. 



The most striking results we observe are (1) that manuring with a.sh con.stituent8 

 alone increa.ses the production of nitrates in the soil ; (2) that the bigger crops grown 

 by ammonium salts or sodium nitrate, with ash constituents, are followed by an 

 increased production of nitrates: Ci) that the use of an organic manure like rape 

 cake or farmyard manure is attended with a large increa.se in the production of 

 nitrate, even aftiT the first active stage of nitrification is long past. * • • 



The manures most readily nitrified by a fertile soil are the salts of annnonia [and 

 it is with these that the phenomenon has been most thoroughly btudied. Th« 



