SOILS FERTILIZERS. 621 



The fertilizing- value of sewage and sewage sludge, H. W. Clark (Mo. Bui. 

 Bd. Health Mass., n. ser., 8 {1918), No. 12, pp. 473-Jt90). — Reviewing experience 

 abroad as well as in this country, the author concludes that only under the most 

 favorable conditions can the returns from sewage farms be made to pay operat- 

 ing expenses, and he maintains that an instance is yet to be cited where these 

 returns pny both the cost of operation and interest on the capital invested. 

 " The exceptions, perhaps, to this are certain tracts or farms in regions of low 

 rainfall and where the sewage is valuable as a liquid, that is, for real irrigating 

 purposes." 



It is stated that the fairly strong domestic sewage of the city of Lawrence 

 contains about 4.28 parts of nitrogen, 1 part of i)hosphoric acid, and 1.5 parts of 

 potash per 100,000 parts of sewage, making the sewage worth as a fertilizer 

 about 6 cts. per 1,000 gal. 



It is stated that average American sewage contains about 2,400 lbs. of sedi- 

 mentable matter in a million gallons, and this sludge contains fertilizing and fatty 

 materials worth approximately $15 to $18. The recovery of the fatty matters 

 and the preparation of the sludge for use as a fertilizer has been attempted in 

 a few places, but the profitableness of the practice is still in doubt. 



" It has also been well proved that the nitrogen, phosphoric acid, etc., present 

 are generally in a less assimilable form than the same bodies in ordinary commer- 

 cial fertilizers. The sludge has value, however, and as the processes of drying, 

 pressing, and fat separation are improved, and also as nitrogen advances in 

 price, as seems inevitable, sewage sludge will become of greater agricultural 

 value than it is at present, especially as the basis of a fertilizer enriched by the 

 addition of potash, phosphates, etc." 



SOILS— FERTILIZERS. 



Field experiments and the interpretation of their results, A. Gr^goire (X. 

 Cong. Internat. Agr. Gand, 1913, Sect. 2, Questio7i 1, pp. IS). — The probable 

 error in field experiments and methods of determining and reducing it are dis- 

 cussed. Reference is made to similar studies by Mercer and Hall, Wood and 

 Stratton, Holtsmark, and Mitscherlich (E. S. R., 27, p. 519), and the results of 

 experiments by Wagner are analyzed with reference to the probable error in 

 them. As a result of such an analysis of 250 tests made at least in triplicate it 

 was found that the probable error in each individual test exceeded 4.7 per 

 cent in 50 cases out of 100, 6.6 per cent in 25 cases, 8.5 per cent in 9, and 10.5 

 per cent in 2. 



The author's general conclusion is that out of 100 field experiments, as a 

 whole, the probable error may exceed 5 per cent in 73 cases, 6 per cent in 64. 

 7 per cent in 52, 8 per cent in 39, 9 per cent in 30, 10 per cent in 18, 11 per 

 cent in 9, and 12 per cent in 4. He shows that results from a single plat are 

 almost always worthless. The probable error is decreased by increasing the 

 number of plats, which should not be less than three. 



With larger field experiments it is very difiicult to keep the probable error 

 below 10 per cent, and generally impossible to keep it below 5 per cent. This 

 indicates that the larger majority of the results of field experiments, as ordi- 

 narily conducted, are not only not worthy of serious consideration but may be a 

 veritable detriment to practical agriculture and discreditable to agronomic 

 science. 



Proposal for an international uniform classification of soils, D. A. Louis 

 (X. Gong. Internat. Agr. Gand, 1913, Sect, 2, Question 8, Commun. D, pp. 5).—. 



