CHEMISTRY. 850 



10 minutes in boiling- water until tlie starch gelatinized. They were 

 then divided into 2 gronps, 2 samples of each substance to be treated 

 without and 2 with diastase. To each of the former 2 cc. of the alum 

 solution and then the copper hydrate (0,3 gm. CuO) were added, heated 

 again for 10 minutes, and allowed to cool. The other set was cooled to 

 Gr)3, 10 cc. of the malt extract (100 gm. malt with ."iOO cc, of water) 

 added, kept for 20 minutes at 05°, tlien after the addition of the alum 

 and copper hydrate as in the other, heated again in boiling water for 

 10 minutes. 



For filtering, the author prefers the folded filter 15 cm. in diameter. 

 Jn some cases mercury was used in the Kjeldahl digestion and then in 

 the distillation fiask enough sodium sulphid was added to precii)itate 

 both mercury and copper. But in most cases no mercury was used and 

 the oxidation was completed by using potassium permanganate at the 

 conclusion of the digestion. 



Corrections must be made for the nitrogen contained in the malt 

 extract and in the filter paper. Tliese were determined by making 

 blank determinations, using the same amount of reagents as were used 

 in the other experim-ints. 



The results of these determinations show that the diastase does not 

 so alter any part of the albuminoid substances as to j)revent their pre- 

 cipitation by the copper hydrate. The amount of albuminoid nitrogen 

 found witli and Avithout the use of malt agreed very closely. 



In the practical carrying out of the analysis it is recommended that 

 the diastase be added after the copper hydrate. This process differs, 

 therefore, from Stutzer's in that after the li(iuid has cooled down to 

 G5° 10 cc. of the malt extract is added and the whole kept at 05° for 

 20 minutes, or is simply allowed to cool. — J. T. A^'DEKsnN, 



The quantitative determination of perchlorate in saltpeter, 

 F. WmTl-:LER (Chem. Zf//., 21 {1897), Xo. JO, 2>p. TTj, ;y;),— The author 

 points out that the method recently proposed by Erck (see below) 

 Involves an ai^preciable loss of chlorin as perchlorate and that where 

 only a little perchlorate is present it may not be detected by the 

 method. He proposes another winch is said to be free from that objec- 

 tion and otherwise reliable. It depends on the fact that fuming nitric 

 acid at a temperature above 200"^ C. is ca])able of reducing the per- 

 chlorate to the chlorid. The clilorate is also reduced, but can readily 

 be removed from a mixture with the i)ei'chlorate by evaporation with 

 concentrated hydrochloric acid. In experiments ,vhich he reports the 

 author used a pure potassium ])('r(;hlorate. A given (juantity of mate- 

 rial in the solid form is placed in a strong glass tube, (!apable of 

 resisting both heat and pressure, silver nitrate is added in excess to 

 precii)itate the chlorin from the chlorid which results from the reduc- 

 ing process, the required amount of fuming nitric acid is poured in, 

 and the tube is hermetically sealed. Alter exposing the tube and its 

 contents to the required temj)erature for 5 hours, the resulting silver 



