978 EXPERIMENT STATION RECORD. 



Attention is especially called to tlie facts that the losses from cov- 

 ered soil were much greater in 1893 than in 1894, and that both the 

 losses and the ratios of drainage to rainfall were higher than those 

 reported by Schlossing (see below). 



The loss of nitrogen in drainage Tvater, T. Schlossing {Compt. 

 Eend., 120 {1895), No. 10, pp. 526-530).— T\\q author maintains that the 

 studies on this subject, made for the most part in vegetation pots or 

 boxes, for obvious reasons, have given results much in excess of the 

 truth, and that examinations of the water of streams draining a given 

 area will furnish a more accurate measure of the loss of nitrogen in 

 drainage. Determinations of nitric nitrogen in the waters of the Seine, 

 Marne, Yonne, and Oise are reported, and estimates based on the results 

 are given which indicate that each hectare of soil in the Seine basin 

 loses 4.2 kg. of nitric nitrogen if one sixth of the rainfall escapes in the 

 drainage, 6.44 kg. if one fourth escapes, and 8.48 kg. if one third 

 escapes. It is stated that while these results do not warrant definite 

 conclusions they nevertheless indicate that the losses of nitrogen in the 

 drainage water are not so important as has been supposed. 



A chemical process of purifying -water, F. Bordas and C. Girard {Compt. Rend., 

 120 (1895), No. 12, pp. 689-691). — In this process slightly more permanganate of lime 

 than is necessary to oxidize the organic matter is added to the water and the excess 

 of permanganate reduced and the color destroyed by manganese dioxid. 



Soil temperatures {Nebraska Sta. Bui. 37, pp. 96, charts 12). — Charts are given 

 showing soil temperatures at the station at depths of from 1 to 36 in. during each 

 mouth of 1894. 



Soil temperatures. R. D. Newton {New York State Sta. lipt. 1893, pp. 734-739).— 

 Tridaily readings of soil thermometers at depths of 1 to 18 in. are recorded for 6 months 

 (May to October) of 1893. 



FERTILIZERS. 



The influence of oxid of iron and alumina in the reversion of 

 superphosphate, A. Smetham [Jour. Soc. Chem. Ind., 14 {1895), No. 2, 

 pp. 112-114; 3, pp. 242, 243). — The chemical changes which take place 

 in the reversion of the monophosphate of lime in superphosphates to 

 dibasic phosphate are explained, and it is claimed that "■where fine 

 grinding is the rule, and where the mixing is complete, and the acid 

 usually in excess of that required for the transforming of the tribasic 

 phosphate to monobasic form, reversion due to the formation of a dibasic 

 phosphate is seldom, if ever, met with." When, however, as is usually 

 the case, the original phosphate contains an appreciable amount of oxid 

 of iron or alumina it is found that the solubility of the phosphoric acid 

 of the superphosphate diminishes to a considerable extent on keeping, 

 any quantity over a very small percentage of oxid of iron and alumina 

 being considered "as reverting double its weight of soluble phosphate.'' 



In studying the relative effects of oxid of iron and alumina in caus- 

 ing reversion the author selected samples of known history and char- 



