SANGER. — VOLATILE COMPOUNDS OP ARSENIC. 139 



which he evidently thinks may take place in this case, giving rise to 

 arseniuretted hydrogen and an arsenical compound of carbon with the 

 hydroxyl group. On the other hand, he refers to the work of 

 Missaghi,* winch opposes the formation of hydrogen by moulds. 



The second paperf of Gosio need be referred to very briefly. The 

 action of P. brevicaule proved so sensitive as to suggest a very delicate 

 method for the detection of arsenic in presence of large amounts of 

 organic matter. The method is carried out as follows. A test-tube 

 is constricted about 20-30 mm. from its lower end, and the bulb thus 

 formed filled with moist cotton wool. A strip of paper is cut in halves 

 and the suspected substance placed between the slices, which are then 

 put into the test-tube and sterilized. Inoculation is then made, and in 

 a day or two the characteristic odor is developed. The temperature is 

 best about 37°. Should further proof be desired, a rubber stopper 

 with right-angled tubes may be inserted in the test-tube, and air 

 drawn slowly through it into a sulphuric acid solution of potassic per- 

 manganate kept at GO to 70°. This solution after proper treatment is 

 introduced into the Marsh flask. 



Undoubtedly in cases where the amount of arsenic is very small, and 

 in contact with a large amount of organic matter, this method would 

 be excellent, but I cannot see that in general medico-legal work it 

 has any advantage over the common methods, particularly as it 

 requires much time and cannot be made quantitative. It is simply a 

 very interesting micro-biological method. The exceedingly great 

 delicacy claimed for it (1 part of sodic arsenite in 1,000,000 being 

 detected in a milk culture) is not surprising when we consider the 

 amount of substance that can be recognized by the sense of smell, 

 E. Fischer and Penzoldt,J for instance, claiming to have detected one 

 four hundred and sixty millionth of a milligram of mercaptan. 



The correctness of Gosio's work is unquestionable, and to him is 

 due the credit of settling this much vexed question. Yet it seemed 

 to me that the importance of the matter demanded a substantiation of 

 his results by repetition of his work, particularly as so many observers 

 beside myself had obtained opposite results, and would be better satis- 

 fied to kuow that the action of the moulds had been confirmed by 

 another investigator in another country. I have therefore made 

 some experiments similar both to those of my first series and to those 



* Gazz. oliim. Ital., V. 419. 



t " Sul Kiconoscimento dell' Ar^enico per Mezzo di alcune Muffe," Roma, 

 1892. 



% Ann d. Pliys. u. d. Chem., CCXXXIX. 131. 



