216 EXPERIMEKT STATION RECORD. 



verting nitrate nitrogen into ammonia nitrogen, viz, Bacillus nii/coides, B. suh- 

 tilifi, B. mesoitericits, and others. 



It was found that Clostridiiiiii gclatinosum in an arabinose medium converted 

 46 per cent of nitrate nitrogen into ammonia nitrogen and utilized about G per 

 cent of the nitrogen in the production of protein substances. 



[Indian soils and fertilizers], .J. W. Leather {Ann. Rpt. Imp. Dept. Agr. 

 [India]. V.X) ',-:>. pp. 53-63, fig. i).— The work of the agricultural chemist to 

 the Government of India during the year ended June 30, 1905, and during 

 previous years is summarized, including studies of available plant food in 

 soils, alkali, combined nitrogen in rain and dew, amount and nitrate content 

 of drainage water, composition of well and canal waters, composition of manures, 

 purification of sewage, composition of canal silt, and methods of artificial 

 preparation of saltpeter. 



The soil studies indicate that nitrogenous manures are generally needed by 

 Indian soils, the sulphates vary widely and are sometimes deficient ; in the usar 

 or alkali soils the pi'oportion of alkali in the upper G or 12 in. rarely exceeds 

 1 to 2 per cent and is commonly much less. The principal constituents of the 

 tilkali are sodium carbonate and bicarbonate, silicate, sulphate and chlorid, and 

 magnesium sulphate and chlorid are occasionally present. The greatest injury 

 is caused I)y the sodium carbonate, both by its corrosive action and by its bad 

 effect on drainage. The most effective remedies for alkali which have been 

 tried are good C'ultivation combined with heavy manuring and the application 

 of gypsum. Drainage has not proved successful on account of the silting up of 

 the tile. The gypsum treatment is too expensive for general use on vei'.v bad 

 land. 



The total nitrogen in the annual rainfall (S7.45 in.) at Dehra Dun was 3.59G 

 lbs. per acre and at Cawnpore (49.30 in.), 3.25 lbs. Dew was much richer in 

 nitrogenous compounds than rain water (2 to 4 parts per million). The pro- 

 portion of nitric nitrogen was relatively larger in dew than in rain. It usually 

 equaled and sometimes exceeded the ammonia, the reverse of what would be 

 expected. Percolation at Cawnpore was the same from a 3-ft. drain gage as 

 from a G-ft. gage. 



The amount of nitric nitrogen ren)oved per year in drainage water from bare 

 soil was about 74 lbs. per acre for the 3-ft. gage and 123 lbs. for the G-ft. gage. 



It is reported that examination of cei'tain well waters from Gujarat, which 

 had been found to have a jiarticularly beneficial effect on tobacco, showed the 

 I)resence of large amounts of potassium, sodium, and calcium nitrates. The 

 proportion of nitrates was so large that the waters could not be used without 

 previous dilution. 



The investigations on sewage reviewed in this article have already been 

 referred to in the Record (E. S. R., 15, p. 7G2). One of the principal objects of 

 these investigations was to ascei'tain the rate of hydrolization of organic nitro- 

 gen in septic tanks and of luirification in contact beds. It was found " that the 

 late of change was considerably greater when the concentration of this organic 

 nitrogen was high than when it was low, and that for a strong sewage a 

 I'elatively smaller tank would be required than for a weak sewage. Secondly, 

 there seemed to be a limit to the destruction of this organic matter. . . . After 

 the amount of this nitrogen had fallen to about 1 part per 100,000 (hundred 

 thousand), no further diminution occurred. . . . Out of every 100 parts of nitro- 

 gen in the fresh sewage, from 10 to 20 were lost in the septic tank, and a fur- 

 ther 10 to 20 were lost in the bacteria bed treatment. The total loss experienced 

 varied from 30 to 40 parts. The magnitude of this loss has not apparently any 

 relation to the strength of the sewage, for it was nearly as great when the 

 weak as when the strong sewage was being treated." 



