AGRICULTURAL CHEMISTRY AGROTECHNY. 807 



per cent of sodium chlorid wns found to have a resistiince of 410 ohms as 

 against a calculated resistance of 420 ohms." 



A modified Kjeldahl flask for determining' soil nitrogen, H. A. Xoyes 

 (Jour. Aiiicr. Clicin. Soc, 36 {1914), Xo. 12, pp. 25.',1, 25.',2, fig. i).— This modified 

 flask is said to do away with the bumpinj; and consequently makes a transfer 

 unnecessaiy. The flask has the same proportions as the oixlinary Kjeldahl 

 bottles, but has a more pointed l)ottom. 



Methods for the chemical, biological, and bacteriological examination of 

 water, O. Emmkrling (Prakdhum dcr Chcmischcn, Biologlschen nnd Bak- 

 teriologischen Wasseruntersuchuug. Berlin: Bonitraeger Bros., 191^, pp. VII-{- 

 200, figs. 171). — ^This work deals with the analysis of drinking water and other 

 waters. In the chapter on the examination of mineral water the determination 

 of radio-activity is included. The biological portion considers the examination 

 of sediments (livinLC and dead matter), while the bacteriological section gives 

 general and special methods for detecting Bacillus coli, B. aiithracis, B. 

 ti/phosus, and the cholera vibrio. The interpretation of results of water analyses 

 is also include<1. 



Hypothetical combinations in water analysis, R. B. Dole (Jour. Indus, 

 and Engin. Chcm., 6 (lOl-i), Xo. 9. pp. 710-714). — This paper discusses the 

 present confusing condition in regard to the reporting of results of water 

 analysis. It shows the advantages of reporting the results in ionic form. 



Methods of estimating carbohydrates. — II, The estimation of starch in 

 plant material. W. A. Davis and A. J. Daisii (Jour. Agr. Sci. [England], 6 

 (1914), Xo. 2, pp. 152-168, figs. 1). — Continuing previous work (E. S. R., 32, 

 p. 112), the authors conclude that "the Sachsse method'' of estimating starch 

 is unreliable in the case of plant material ; not only does the presence of 

 pentosans falsify the results, as pentoses are formed during the hydrolysis, 

 but actual destruction of dextrose occurs during the prolonged treatment with 

 dilute acid. O'Sullivan's method gives low results, owing to the loss of dex- 

 trin which occurs during the purification of the solution after the conversion 

 by diasta.se."' 



"To estimate starch, the dry material (free from sugars and, if necessary, 

 previously extracted with water to remove gums, amylans, etc.) is gelatinized 

 with 200 cc. of water in a beaker flash heated for one-half hour in a water bath 

 at 100° C. The solution is cooled to 38", 0.1 gm. taka-diastase added, together 

 with 2 cc. of toluene, and the mixture left 24 hours in order that the conversion 

 may take place; it is then heated in a boiling water bath to destroy the diastase 

 and the clear solution above the residual leaf material is filtered through a 

 fluted filter paper into a 500-cc. measuring flask; the leaf residue is thoroughly 

 washed several times by decantation, the washings being passed through the 

 filter paper until the volume of liquid in the flask amounts to aiwut 475 cc. 

 The necessaiy quantity of basic lead acetate is then adde<l to precipitate the 

 tannins, etc., present in the solution; the amount required varies considerably 

 with different leaves, generally ranging from 5 to 25 cc. A large excess of lead 

 should be avoided and tests should be made after each small addition of lead 

 acetate in order to ascertain when the precipitation is complete. When this is 

 the case, the solution is made up to 500 cc. at 15° and filtered; 100 cc. of the 

 filtrate is placed in a 110-cc. measuring flask, the slight excess of lead precipi- 

 tated by adding solid sodium carbonate and the volume adjusted to 110 cc. at 

 15°. Fifty cc. of the filtrate from the lead carbonate is used for the reduction 

 and another portion polarized in a 400-mm. tube. 



The method of calculation is explained. 



•Sitzber. Naturf. Gesell. Leipzig, 5-7 (1877), pp. 30-^37. 



