AGRICULTUR-^ CHEMISTRY. 1111 



for commercial and factory work. The original (iray method gives results 

 which are mucli too low (as much as 1.0 per cent in sample 20) to be reliable," 



A further comparison of methods at this institution is noted on page 1178 of 

 this issue. 



The caprylic acid determination in butter fat, R. K. Dons (Ztschr. Unter- 

 siicli. Xtihr. u. Genussmil., 15 (1908), Xo. 2, pp. 75-79). — The method proposed 

 depends upon the fact that caprylic acid is only slightly soluble in water but 

 is easily soluble in a mixture of myristic, palmitic, and oleic acids. After the 

 usual saponification of the butter fat and decomposition of the soap, shaking 

 with water at 80° C. removes the soluble volatile acids, as butyric and caproic, 

 and the caprylic acid is then separated from the insoluble acids by distillation 

 and titrated with silver nitrate. • 



Estimation of cocoanut oil in butter, R. Ross and J. Race {Chern. News, 97 

 (1908), Xo. 2519, pp. 110, 111; ahs. in Analyst, 33 (1908), No. 385, p. 122).— 

 The authors consider the process described by Hodgson (E. S. R., 19, p. 014) to 

 be useless. They point out that the quantity of sulphuric acid proposed has 

 a considerable action on the permanganate used for oxidation. With smaller 

 quantities of acid the oxygen equivalents obtained by oxidizing fats are unsatis- 

 factory for differentiating one fat from another. 



The Polenske number, M Fbitzsche (Ztschr. Untersuch. Xalti: u. Genuss- 

 mtl., 15 (1908), Xo, Jf, pp. 193-233). — The experiences of other investigators 

 in the determination of the Polenske number are reviewed, especially with re- 

 gard to the method of making the test for the estimation of cocoa fat in butter 

 and to the judging of butter on the basis of this constant. The author believes 

 that the Polenske method is on a sound basis, and that the various modifica- 

 tions that have been suggested offer no substantial advantage. The quantity 

 and fineness of the pumice-stone powder used in the distillation are important. 

 Satisfactory results are obtained when, with strict adherence to the other con- 

 ditions of the test, 0.5 gm. of the powder of 1 mm. fineness is used. 



The important results of the author's w^ork in the determination of the Po- 

 lenske number in 355 samples of pure Holland butter are reported and dis- 

 cussed. In the majority of the tests the value was below the minimum given 

 by Polenske, and in none was the maximum exceeded. In butter from most 

 of the Netherlands provinces the constant was lowest in March and April and 

 highest in Fel)ruary. In no case was the results affected l)y the feeding of 

 cocoa cake or beets. 



Distillation apparatus for nitrogen determinations, C. Blocii (I'Ji<nin. 

 Z'f(/.. 52 (1907), p. 311 ; <ihs. in Ztschr. Anycw. Chcm., 21 (1908). Xo. 7. p. 

 315). — The special feature of this api)aratus is a funnel attached to the dis- 

 tillation flask so that solutions may be added without opening the fiask. 



Passerini's calcimeter, P. Fanteciii (liol. 1st. Af/r. t^cimdicci, 2. scr., 7 

 (1908), Xo. 1, pp. 103-107, flg. 1). — The construction and method of using this 

 pie<'e of apparatus for the quick determination of the lime content of soils art- 

 described. 



An accurate form of gas analysis apparatus for commercial and other 

 purposes, W. A. Rone and R. V. Whkelek (Jour. Soc. Chcm. Indus., 27 (1908), 

 Xo. 1, pp. 10-12, flg. 1). — The authors describe '" a convenient and not too costly 

 adaptation of the gas analysis apparatus used throughout their researches on 

 c<miI)ustion, which shall meet the daily requirenu'uts of a conunercial research 

 laboratory." 



Regulations for the examination of feeding stuffs in the laboratory of 

 the Colonial Museum at Haarlem {Hcpnratr from I'liarm. Wcckbl., .} J (1907), 

 Xo. 23, ISup., pp. 29). — A second edition of the Dutch otlicial methods for the 



