1917] AGRICULTURAL CHEMISTRY — AGROTECHNY. 715 



the recently admitted tartrazin, is described. A scheme for the separation of 

 the dyes is included. 



The determination of butter fat in margarins, G. Jorgensen {Ann. Falsif., 

 9 {1916), No. 94-95, pp. 262-27 0) .—On account of the large amount of caproic 

 and caprylic acids in palm and coconut oil the author considers the Reichert- 

 Meissl number unreliable as an index of the amount of butter fat in margarins. 

 The following procedure, in which the water-insoluble silver capronate and 

 caprylate are precipitated as earlier suggested by Kirschner (B. S. R., 16, p. 

 1050) and Jensen (E. S. R., 17, p. 697), is described: 



To the neutral liquid from the Reichert-Meissl determination 30 cc. of tenth- 

 normal silver nitrate is added. The mixture is then made up to 150 cc, 

 thoroughly agitated, set aside for one hour, and filtered through a folded filter. 

 To 100 cc. of the filtered liquid 20 cc. tenth-normal sodium chlorid is added, and 

 the excess sodium chlorid then titrated with standard silver nitrate, using 

 potassium chromate as indicator. The number of cubic centimeters of silver 

 nitrate used in the titration multiplied by 1.5 gives the amount of insoluble 

 silver salts {A), corresponding to the soluble volatile acids. Designating the 

 Reichert-Meissl number as B, the silver index is then equal to B-A. 



Having the silver index and either the Polenske number or the total volatile 

 acids, the percentage of butter fat in the sample is easily calculated by formulas 

 which are submitted and described. 



The method is deemed to yield results sufficiently accurate for practical work. 

 A table giving the percentages of butter fat equivalent to the silver indexes 

 obtained, together with the amounts of total volatile acids and Polenske num- 

 bers, is Included. 



A boiling method for the determination of water-soluble arsenic in lead 

 arsenate, G. P. Gray and A. W. Christie {Jour. Indus, and Engin. Chem., 8 

 {1916), No. 12, pp. 1109-1 US). —The authors, at the California Experiment Sta- 

 tion, propose the following procedure : 



Five-tenths gm. of the dry powder is placed in a 500 cc. Erlenmeyer flask, and 

 after the addition of 200 cc. of distilled water boiled briskly for 10 minutes. 

 If the liquid is cloudy or the material does not settle readily, it should be 

 further digested on the steam bath for an hour or two. The material is filtered, 

 washed with a small quantity of hot water, and the filtrate collected in a 500 

 cc. Erlenmeyer flask. One gm. potassium iodid and 4 cc. of concentrated 

 sulphuric acid are added and the volume of the liquid reduced to 40 cc. by 

 boiling. The liquid is then diluted to 200 cc, and the free iodin remaining 

 titrated with approximately twentieth-normal sodium thiosulphate. The solu- 

 tion is nearly neutralized with concentrated sodium hydroxid, using methyl 

 orange. Sodium bicarbonate is added to alkalinity and then to excess. The 

 arsenic is titrated with a twentieth-normal iodin solution, starch paste being 

 used as indicator. 



The procedure outlined is considered to be shorter than any recommended 

 heretofore, only three to four hours being required for an analysis. It yields 

 accurate results, and its use is inexpensive. 



The results obtained by the new procedure are higher than those obtained by 

 the provisional method of the Association of Ofiicial Agi-icultural Chemists, but 

 are deemed to be more nearly correct. 



The methods in use at present are briefly reviewed and discussed. 



An oxalate-iodid process for Paris green analysis, C. A. Peters and L. E. 

 Fielding {Jour. Indus, and Engin. Chem., 8 {1916), No. 12, pp. 1114, 1115). — A 

 method for the analysis of Paris green, in which the copper is precipitated as 

 oxalate, the combined oxalic acid titrated with permanganate, and the arsenic 

 oxidized in the filtrate by iodin, is described in detail from the Massachusetts 



