918 



of tlie sulphates contained in soils, particularly of sulphate of liiiic* 

 it was thou<;ht to l>e of interest to study tlie conipanitive action 

 in this respect of sulphate of iron and sulphate of lime iucorporated 

 in various artificial soils of known composition. + For this pur- 

 pose artificial soils were prepared as follows: (1) Cotton-seed meal 

 (containing 4.01 per cent of nitrogen) 0.04 kg., and pure siliceous 

 sand 1.96 kg.; (2) cotton-seed meal 0.04 kg., sand 1.70 kg., and clay 

 0.2 kg.; (3) cotton-seed meal 0.04 kg., sand 1.94 kg., and carbonate 

 of lime 0.(^2 kg. ; (4) sand 1.74 kg., cotton-seed meal 0.04, carl)onate 

 of lime 0.02, and clay 0.2. These mixtures were i>laced in enameled 

 earthenware pots ^10 in all), in the bottoms of which had been pre- 

 viously placed fragments of glass, forming a space through which air 

 could circulate, and from which large tubes ])assed up through the soil 

 and communicated with the air. These tubes served not only for adnnt- 

 tiiig air but also for introducing the water used in moistening the soil. 

 Difterent pots containing these mixtures received sulphati' of iron 

 (FeS()4J at rates of 1, 2, and '.i grams per kg. of soil, ami suli»hate of 

 lime at the rate of 5 grams per kg. In addition, jmts containing 

 mixture No. 4 received, in difierent <-ases, lactate of iron at a rate cor- 

 responding to 1 gram of sulphate of iron per kg., and ferric oxide 

 at a rate of 2 grams jter kg. The pots were inoculated with 10 c. c. of 

 water-extiact of nitrifying soil and pure water was added from time to 

 time. The experiments lasted from ^I;iy 2 tt» DeceMilx'r 12, at which 

 date the contents of the pots were removed and carefully analyzetl. 

 Full details regarding gain or loss of the various i'orms of nitrogen 

 in each of the 16 experiments are tabulated and discussed at length. 



In pure sand the addition of 1 giam of sulphate of iron to 1 kg. 

 reduced the loss of nitrogen from 47.»1."> to 1S,:{0 ]>er cent: at the 

 same time the gain of nitric nitrogen was increased from l.l.'J to 10.4 jter 

 cent, and of ammoniacal m'trogen from 4.49 to 11.22 per cent. NN'lien 

 clay was iidded to the soil at a rate of 10 grams to 1 kg. of soil the 

 sulphate of iron sensibly jcduced the loss of nitiogen. but tin* gain of 

 nitric nitrogen was increased from 5.1 to 1."».92 i>er cent, ami of ammo 

 niacal nitiogen reduced from 1().94 to I2.7."» per cent. The addition <if 

 carbonate of lime to the same mi>8tnre did not sensibly diminish the 

 loss of nitrogen; the gain of nitric nitrogen was diminisluMl trom ."i.l 

 to 2.55 per cent, and that of ammoniacal nitrogen was increased from 

 C.4;{ to 10.4 i>er cent. 



In the presence of <'arbonate of lime and clay the sulphate of iron 

 reduced the loss of nitrogen from 31.03 to 23.87 per cent aiwl diminished 

 the lu'odiiction of nitric nitiogen ti'om 1S.57 to r).43 per ••cut. In tiie 

 same medium double and triple doses of suli)hate of iron reduced the 



•Compt. rend., May 16, 1884, and Soptenibor 9, 1889. 



t Ktueiit experinuMits on the eft'cct on nitritication, etc., of diUVrcnt ]iid)iortion3 

 of clay and organic nitrogen were repoitt-d in Compt, rend., lU (18!ll.'), pp. 81-84 

 (see Experiment Station Kecord, vol. iii, p. {'>M\), 



