420 F. A. Per ret — Vesuvius. 



Nitroprussiate soda 3 parts 



Zinc chloride, fused _ 5 parts 



Ferrocyanide potassium 3-4 parts 



The precipitate is of a light yellow color, and is not affected 

 by light. Spread on filter paper and moistened with dilute 

 ammonia, it turns to a reddish purple when exposed to S0 2 . 

 It has the great advantage of not being affected by hydro- 

 chloric or hydrofluoric acids, both of which are so often pres- 

 ent in fumarolic exhalations. 



For the detection of hydrochloric acid a glass rod dipped in 

 a solution of silver nitrate is exposed to the gases and immersed 

 in dilute nitric acid, when, if HC1 is present, the silver chlor- 

 ide formed will be precipitated. As an alternative, a strong 

 solution of ammonia on a glass rod may be exposed to the 

 gases with the formation of white vapors of ammonium chlor- 

 ide by hydrochloric acid, but the former test is more delicate 

 and reliable, especially in a strong wind. 



For laboratory analyses the gases are collected in expanded 

 glass tubes, which are then sealed off with a benzine pressure 

 lamp. It is generally necessary to aspirate the gases through 

 the collecting tubes and for this a rotary tube-pump may be 

 employed, but occasionally there is sufficient pressure to drive 

 the gases through the tubes and through a water-valve at the 

 farther end, provided the collecting tube is connected to a 

 funnel placed over the fumarole and banked around with 

 earth. Bunsen used tubes of tin for insertion in the fumaroles, 

 but the temperature of some of the present Vesuvius vents 

 precludes their employment. I have found no difficulty in 

 using glass provided the tube is bent downward just above the 

 fumarole, otherwise water condenses above and trickles down 

 to the hot portion, causing the tube to crack. 



In the general investigation of fumaroles it is important 

 to note that those which develop on lava streams at a distance 

 from their mouth of exit have no connection with the interior 

 of the mountain, and cannot, therefore, serve as indicators of 

 its condition. They are formed in connection with fractures 

 in the lava stream and have an evanescent existence, which is 

 doomed to extinction with the progressive cooling of the lava. 

 These often act, for a time, as true fumaroles, bringing forth 

 and depositing the volatilized products of the lava stream, but 

 they soon degenerate into carriers of perfectly neutral hot air, 

 with which, after rain, is mingled the vapor of water. They 

 are not, therefore, true fumaroles of the volcano but fumaroles 

 of the superficial lava stream, and are of the secondary type as 

 contrasted with those which form in connection with fissures 

 in the mountain, and which may be expected to rise in temper- 

 ature with its increasing internal activity. 



