112 EXPERIMENT STATION RECORD. [Vol. 40 



A study of sources of error incident to the Lindo-Gladding method for 

 determining potash, T. E. Keitt and H. E. Shive;: (Jour. Indus, ami Ennin. 

 Chem., 10 (1918), No. 12. pp. 99.)-99G). — A critical examination of the Lindo- 

 Gladding method for determining potash is reported from the South Caro- 

 lina Experiment Station, which indicates that there are two sources of error in 

 this method: (I) The volume of the solution is decreased by the bulk ol the 

 precipitate formed on addition of ammonia and ammonium oxalate, which 

 makes a plus error, and (2) the potash in solution is decreased by occlusion of 

 potash by the heavy gelatinous precipitate formed. These two sources of error 

 are partially compensating. 



It was found impossible to wash out with hot water the potash occluded 

 within the precipitates. It may, however, be separated to a certain extent by 

 repeatedly dissolving the precipitate in hydrochloric add, diluting to a large 

 volume, precipitating With ammonia and ammonium oxalate, filtering, and de- 

 termining the potash in the filtrates and washings. Both Iron and calcium 

 phosphate when precipitated with ammonia were found to occlude potash. 



The quantitative determination of phosphdrus by the nephelometric 

 method, E. B. Meigs (Jour. Biol. (hun.. 36 [1918), No. ..'. pp. 335-846).— The 

 author has made a study at the Bureau of Chemistry of the U. 8. Department 

 of Agriculture, of sources of error in the nephleometric determination of phos- 

 phorus, the roults Of which are summarized as follows: 



"The greater the concentration of hydrochloric acid present when the 

 strychnin sulphate is added, the greater will be the amount of material pre- 

 cipitated at that time and the greater will he the stability id the reagent. 

 Reagents made up with strong hydrochloric acid, however, tend to minimize 

 the differences in nephelometric value as between phosphate solutions of differ- 

 ent densities. None of the reagents that arc serviceable when used in the 

 procedure described by I'.ioor [in the article previously noted (B. s. EL, 35, p. 

 1GG) 1 are entirely stable; they all tend to form a spontaneous precipitate and 

 to reach a stale in which they show no difference in nephelometric value as 

 between phosphate suspensions of different densities. The temperature at 

 which the strychnin sulphate is added to the Na. MoO«-HGl solution in making 

 up the re: l'i lit may have a < siderable effect on Its character." 



The temperature at which the phosphate suspensions are precipitated is also 

 considered to be a factor which may cause a lack of proportionality between 

 differences in the densities of suspensions and the differences in their nephelo- 

 metric values. 



Determination of alkaline carbonates and bicarbonates in the cold in the 

 presence of litmus and phenolphthalein. Analysis of a mixture of carbo- 

 nates. W. Mi.snazAT (Ann. Chim. Analut., 23 {1918), No. 10, pp. 201-206).— 

 Certain modifications in technique are described, by means of which it is said 

 to be possible to analyze with accuracy a mixture of alkaline carbonates and 

 bicarbonates by titration in the cold, using litmus and phenolphthalein as in- 

 dicators. 



To determine bicarbonates with accuracy it is recommended that the solution 



be diluted with boiled water to bring the proportion of alkaline carbonates to a 



figure below 0.07 per cent calculated as sodium carbonate. In titrating with 



N 

 litmus, it is suggested that when the addition of "- acid to the alkaline liquid 



begins to produce a color change the liquid should be divided Into two parts, 

 one nf which receives the acid and the other serves as a comparison. If the 

 acid added brings about a change in color the operation is continued, the 

 contents of the two flasks being mixed and then again divided into two portions. 



