226 Proceedings of the Royal Irish Academy, 



teristic properties, may be easily recogiiised, and thas very minute 

 quantities of arsenic under different cii'cumstances can be readily 

 detected. This method, as proposed by its discoverer, consists in gene- 

 rating, in a suitable apparatus, hydrogen by the action of dilute 

 sulphuric acid on metallic zinc, and then adding in the state of solution 

 the arsenical compound, when arseniuretted hydi-ogen will be quickly 

 generated, and a fine jet of the gas being ignited, and a cold surface 

 placed down on the top of the flame, very characteristic spots or stains 

 of metallic arsenic will be produced ; or the gas being passed thi'ough a 

 heated tube, it will be decomposed, and a metallic sublimate formed at 

 a short distance beyond the heated portion. I need not refer to the 

 apparatus recommended by Mr. Marsh for carrying out his test, as it is 

 now so well known, nor to the modifications of it which have been 

 subsequently proposed ; and I must acknowledge that this beautiful 

 means of detecting arsenic, owing to its great delicacy and veiy conclu- 

 sive results in the hands of the experienced chemist, leaves but little 

 to be desii'ed. It, however, labours under this serious disadvantage, 

 that the acid and the zinc which are employed in the process may one 

 or other of them, or even both, contain more or less of arsenic as an 

 impurity, and consequently the indications of that substance which are 

 thus obtained, may be due not to its existing in the suspected matter 

 or object under investigation, but to its occurring as an impurity in the 

 materials employed in this process for its detection; and I may add that 

 it is difficult to get in commerce the zinc and sulphuiic acid required 

 perfectly free from arsenic. 



To obviate more or less this source of fallacy, several modifications 

 of the original process of Marsh have been suggested. Thus, Pleitmann, 

 some years ago, proposed the use of a strong solution of caustic potash, 

 assisted by heat, instead of the acid, to act on the zinc as a means of 

 generating the hydrogen gas, and in this way one source of arsenical 

 contamination was avoided. It was found, however, to be too slow a 

 means of generating hydi'ogen to detect arsenic in the usual way by 

 Marsh's method. Professor Bloxam has suggested the employment of a 

 galvanic battery for the generation of the same gas, and in this way 

 obviates the use of zinc, and thus excludes another possible source of fal- 

 lacy ; but, owing to the trouble and expense attendant on the use of a gal- 

 vanic battery, wHch for this pui'pose must be of some power, and the 

 arrangement being of rather a complicated character, and still requiiing 

 sulphuric acid, it has, I believe, been but little employed. I should 

 also add that the metal aluminium, and more recently magnesium, have 

 been proposed as substitutes for zinc in Marsh's process or in Fleitmann's 

 modification of it, as being less likely to be contaminated with arsenic 

 than that metal. The modification which I would now suggest, and 

 which, as far as I can ascertain, has not hitherto been proposed, is the 

 employment of an amalgam of sodium andmerciuy as a means of gene- 

 rating the hydrogen required for the test : and by the use of this 

 Bubstance I do away with, altogether, the necessity of any acid, and I 

 employ two metals which are not liable to arsenical contamination. As 



