SCIENCE BULLETIN, No. 4. 



The following modification of Marsh's apparatus, as used by the author, is 

 described for the information of those who wish to avoid some of the 

 disadvantages attending the older forms : 



Impure hydrogen, generated from dilute sulphuric acid and ordinary 

 granulated commercial zinc, is purified by passing through a neutral aqueous 

 solution of silver nitrate, which combines with the AsH 3 with reduction to 

 metallic silver. It then passes through a mixture of the suspected liquid 

 with dilute sulphuric acid, in contact with a small quantity of pure arsenic- 

 free zinc, or a bundle of magnesium wire (in which latter case the solution 

 must contain only very little free acid, so as to avoid much heating), and 

 carries over the small quantity of hydrogen thus produced, with the AsHj 

 resulting from the reduction of any arsenic present in the suspected matter. 

 The combined gases, after drying, are tested in the usual way. 



The arrangement described and figured below will be found a simple one, 

 and has given every satisfaction in the author's hands. 



The hydrogen-generating vessel (a) is an ordinary 40-oz. flask, fitted with 

 a safety thistle funnel (g), with side tube, and charged with about 4 oz. 

 commercial granulated zinc. The side tube is connected by a small piece of 

 rubber tubing to a bent glass tube passing through a cork fitting into the 

 neck of a small conical filtering flask (6), containing about 200 cc. of a solu- 

 tion of silver nitrate in distilled water (about 5 per cent.). The side tube of 

 this flask connects with the inlet tube of a second similar filtering flask, 

 containing a further quantity of the same solution. The hydrogen issuing 

 from the second AgNO 3 solution will be found perfectly free from arsenic. 



The purified gas now passes into a flask (c), holding about 400 cc., the 

 rubber stopper of which has three perforations one for the inlet tube leading 

 to the bottom of the flask, another for a stoppered funnel tube (f), also 

 leading to the bottom, and a third for an exit tube. 



A JJ-tube (c?), of calcium chloride, is interposed as usual in front of the 

 constricted hard-glass tube (e), fitted with platinum jet. The flask (c) contains 

 either a few grams of arsenic-free zinc or a bunch of magnesium wire, along 

 with about 20 cc. of water, acidulated with pure H 2 SO4. 



To use the apparatus, pour about 50 cc. of 5 per cent, sulphuric acid on to 

 the zinc in the generating flask ; then after a couple of minutes the burner 

 under the reduction tube (e), which is protected with gauze, is lighted, as well 

 as the gas issuing from the platinum jet. 



More acid should be added from time to time, so as to keep the hydrogen 

 flame about 1 cm. in length. If after an hour no mirror is formed, the 

 experiment is again started with further quantities of the same materials, with 

 the exception of the AgNO 3 solution, which lasts for a large number of 

 experiments, the stoppered funnel tube (f) being filled with the suspected 

 liquid, which is allowed to flow very slowly into the flask (c) containing the 

 pure zinc or magnesium. 



If arsenic is present, it usually commences to deposit in a few minutes. It 

 is advisable, however, in order to detect very minute quantities, to allow the 

 gas to run a full hour. 



