AGRICULTURAL CHEMISTRY AGROTECHNY, 609 



egg which has been treated with hydrochloric acid and neutralized or serum 

 which has been treated in the same manner. The explanation given is that 

 these substances are supposed to contain some of the same materials as rennet 

 and apparently displace it in the charcoal. 



The coloring- matter of the tomato, R. Willstatter and H. H. Escher 

 (Ztschr. I'lnjsloJ. Clicm., U'l {1910), No. i, pp. 47-61, pi. 1, fig. i).— Lycopin, 

 the colorins matter obtained from the tomato by the author, differed markedly 

 in shade from carotin. While these hydrocarbons are alike as to molecular 

 weight and behavior toward oxygen, it is shown that the rate of autoxidation 

 is greater with lycopin, and their combinations with the halogens and their 

 crystallization on cooling the solutions seem to show that they are different 

 bodies. 



Composition of gas from cottonwood trees, F. W. Bushong (Trans. Kans. 

 Acad. HcL. 21 {I'JOl-S), pt. 1, p. .jJ).— Having noticed that gas bubbles in the 

 saj) upon the freshly cut trunk, stump, and chips of cottonwood trees contained 

 an inflammable gas, the author collected gas from a hole bored into the 'heart 

 of a cottonwood tree and submits the results of analyses of it made by 

 McFarland. 



The constituents found were: Oxygen, 1.24; carbon dioxid, 7.21; methane, 

 60.90 ; and nitrogen, etc., by difference, 30.65 per cent, defines, carbon 

 monoxid, hydrogen, and ethane were not found. 



Bacterial action the cause of the corrosion of steel, R. H. Gaines (Abs. in 

 Chciii. /All.. 33 (1!)0!)), No. ]',5, p. 126S). — It is concluded that bacteria are 

 one of the chief causes of corrosion of steel in the soil, as the author found on 

 analyzing the rust that it contained much organic matter and from 1.41 to 

 0.95 per cent of combined sulphur (calculated as SOj), whereas the original 

 steel contained but 0.05 per cent of sulphur. 



Quantitative estimation of the decomposition of potassium nitrate by 

 micro-organisms, H. Franzen and E. Lohmann (Ztschr. PhysioL Clicm., 63 

 (lUOH). No, 1, pp. 53-102; abs. in Analyst, 35 (1910), No. J,06, p. 2//).— Known 

 amounts of iiotassium nitrate were added to peptone media and the tubes 

 inoculated with various bacteria. The amount of nitrate remaining at the 

 end of specified periods, together with the nitrite (after the latter had been 

 oxidized by hydrogen peroxid), was estimated by Busch's nitron method 

 (E. S. R., 16, p. 045). The nitrate alone was also determined after destroy- 

 ing the nitrite with hydrazin sulphate. In this way a measure could be ob- 

 tained as to the amount of nitrite formed. 



On the basis of these experiments the authors classify the bacteria into 3 

 groups, (1) those which convert the nitrate into nitrite without further de- 

 composition of the nitrite (Bacillus plymotithensis, B. prodigiosus, B. kiliense, 

 Proteus vulgaris, B. coli commicnis, and B. typhi murium), (2) those which 

 decompose the nitrate and immediately convert the nitrite into nonoxidized 

 nitrogenous compounds (B. pyocyaneus), and (3) those bacteria which have 

 no effect on nitrates at all (B. fluorescens liquefaciens). 



Determination of small quantities of nitrogen, T. Zeller (Landiv. Vers. 

 Stat., 11 (1909), No. 6, pp. 437 -UO) .—This is a criticism of Mitscherlich's 

 method (E. S. R., 21, p. 208). Comparisons are made between the figures actu- 

 ally found by Mitscherlich and those set down as the limit of error, attention 

 is drawn to the fact that errors are liable to occur in the method, and finally 

 the author raises the question whether it is really necessary to have so accurate 

 a method as this. 



Nitrogen determination in soil extracts, Densch (Chcm. Ztg.. 33 (1909), 

 No. 143, pp. 1249-I25I). — The author has reinvestigated his own method and 

 claims that the error is never higher than 0.3 mg. Mitscherlich's assertion 



