806 EXPERIMENT STATION RECORD. [Vol.43 



data obtained on 100 representative samples of milk, indicating that the volume 

 of the fat and protein in the 66 gm. of milk used as a sample in the Official 

 method for the polariraetric determination of lactose is invariably higher than 

 the correction of 2.6 cc. allowed for every sample. Calculations of the actual 

 volume of precipitated milk fat and protein exceeded the correction of 2.6 cc. 

 by amounts ranging from 0.6 to 5.53 cc, resulting in a discrepancy in the final 

 lactose reading of from 0.03 to 0.28 per cent. 



To overcome this inaccuracy a procedure is suggested which is based upon 

 previous observations of the author (E. S. R., 36, p. 374) and of Huecker 

 (E. S. R., 31, p. 670) that the protein content of normal milk can be estimated 

 with a considerable degree of accuracy when the fat content is known. Tables 

 are given in which the volume of fat and protein in 66 gm. of milk of varying 

 percentages of fat and protein has been computed from the specific volumes 

 of the milk fat and milk protein. This volume represents the correction for all 

 samples of milk whose fat content places them within the group. This correc- 

 tion can be applied equally well by increasing the dilution of the sample to the 

 proper amount, or by using the figure as V in the following formula : 



R 100-V , , , ,., , 



Percentage lactose=^X — -inn , where the samples are diluted to a uniform 



volume of 100 cc. 



Tests of the accuracy of this method are reported, indicating with the rep- 

 resentative samples used a maximum error of 1.46 cc. This amounts to a final 

 error of 0.07 per cent, as compared with a maximum error of 5.53 cc. or 0.28 

 per cent in case of the Official method. 



lodometric method for the determination of reducing sugars, H. van 

 NouHUiJS (Arch. Suikerindus. Nederland. Indie. 28 (1920), No. 11, pp. 591-595; 

 nbs. in Intematl. Sugar Jour., 22 (1920), No. 259, pp. 411. ^12). — Detailed direc- 

 tions are given for the volumetric determination of reducing sugars in molas- 

 ses by estimating the copper remaining in the unreduced Fehling's solution by 

 adding potassium iodid, acidifying, and titrating the liberated iodin with 

 standard thiosulphate. 



A comparison of this method with the volumetric permanganate method is 

 reported which indicates that the results by the iodometric method run a 

 little low. The method is recommended, however, as being very rapid and 

 convenient in practice. 



A picnometer for the determination of density of molasses, W. B. New- 

 kirk (U. S. Dept. Com., Bur. Standards Technol. Paper 161 (1920), pp. 7, fig. 1; 

 also in Sugar [New York], 22 (1920), No. 9, pp. 508, 509, fig. 1).—Xn improved 

 picnometer for determining the density of molasses and other viscous liquids is 

 described and illustrated by a diagram. 



The apparatus consists of a bottle with an enlarged rim ground optically flat 

 and closed with a disk, also gi'ound optically flat. When the bottle is to be 

 filled, the disk is removed and replaced by a spherical expansion chamber 

 which is fitted with a vacuum connection. The molasses to be analyzed is 

 poured into the bottle and expansion chamber until the latter is about one- 

 third full. The vacuum connection is then made and the pressure reduced 

 until the gas expands to visible bubbles. The vacuum connection fs then closed 

 and the apparatus placed in a thermostat until all the bubbles have collected 

 in the expansion chamber and the temperature has reached etiuilibrium, after 

 which the expansion chamber is removed, the bottle closed by the disk, wiped, 

 and weighed. The densities are determined by correcting the weights to vacuo 

 and comparing with the weight of an equal volume of water at 4° C. in vacuo. 

 The results are reported in degrees Brix. 



