714 EXPEEIMENT STATION RKC'OKD. 



The peix'outagc of pota.sh is calculated by nmltiplyiiig the corrected 

 number of cubic centimeters of lead nitrate b>' 1,99. 



Thephospho-molybdic sohition used is prepared by dissolving 140 gm. 

 of sodium carl )oiiatc(Xa2 CO Jand 20 gm. of sodium phosphate (Na„HPO,) 

 in 500 to t)00 cc. of water, adding 70 gm. of freshly ignited moh'})dic 

 acid (M0O3), 200 cc. of nitric acid, making the volume to 1 liter, allow- 

 ing to stand 24 hours, and filtering. This reagent gives with potas- 

 sium, annnonium, and thallium salts as well as alkaloids, a yellow pre- 

 cipitate soluble in ammonia and insoluble in nitric acid. The molyb- 

 denum content of this precipitate varies, but the ratio of potash to 

 phosphoric acid remains constant at 1.00. 



A short method for determining potash in potash salts, H. ]Seu- 

 BAUER {Zt^chr. Analyt. Chern., 39 {1900), No. S, pp. J^l-oOi^.—X 

 method for the estimation of potash in crude salts containing mixtures 

 of sulphates and chlorids of potassium, sodium, magnesium and calcimn. 

 To a 25 cc. solution of 0.5 gm. of substance a few drops of hydro- 

 chloric acid are added and enough platinic chlorid for precipitating 

 the potash as double chlorid salts. The solution is evaporated to dry- 

 ness in a water bath, moistened with about 1 cc. of water, and rubbed 

 with a glass rod with a flattened end. About 30 cc. of alcohol are 

 added in successive portions of 10 cc. and the mass rubbed after each 

 addition. The material is covered and allowed to stand half an hour, 

 receiving an occasional rubbing during that time. It is then trans- 

 ferred to a Gooch crucible and washed with ether. The crucil)le is 

 then gently heated while a current of hydrogen gas is introduced 

 through the lid. After 5 minutes the flame is increased to a dull red 

 heat and continued for 20 minutes. The residue in the crucible is 

 first moistened with cold water, then washed about 15 times with hot 

 water, and finally with a 5 per cent nitric acid solution, without the 

 use of a filter pump, for about 30 minutes. The washing should Ije 

 thorough in order to remove the salts. After this the residue is 

 washed with hot water, ignited, and weighed. The factor 0.48108-is 

 used for estimating the K.,0. 



On cellulose determination, C. Councler {(linn. Ztg., 2J^ (1900), 

 JVo. So, pp. oOH,.jGO). — This is a comparison of the Schulze-Henneberg, 

 MiiUer, and Lange methods, and a modification of the Miiller method 

 proposed by the author. In the Midler method the dried sul)stance 

 is extracted with a mixture of strong alcohol and benzol, and then 

 with hot water. The residue is treated with bromin water and then 

 with dilute ammonia, heating nearly to boiling. This treatment with 

 bromin water and animonia is repeated a number of times until the 

 material is not colored when heated with dilute ammonia. It is then 

 washed with water and with boiling alcohol and dried at 110 to 115° C. 

 "While Midler claims that 3 or 4 treatments with l)romin and ammonia 

 are suthcient, the author finds that sometimes as manv as 20 are 



