Analysts of a singular suhstaiice Jbund at Vesuvius. 379 



some places, where it appeared to be thickest, the colour exactly 

 resembled that of a dark Seville orange, while in others it had 

 a more yellow hue, and, when taken up, appeared soft and 

 claggy." 



When I received the bottle, containing the substance in ques- 

 tion, in September 1831, about a year after it had been collect- 

 ed, it presented the appearance of a mass of the consistence of 

 butter, and of a bright orange colour, in some parts clean and 

 pure, in others mixed with a considerable quantity of a dark 

 sand. The bottle, when opened, had a smell which led me tp 

 suspect the presence of bromine or iodine ; but in the course 

 of a few hours this smell was gone, and I could not detect the 

 presence of either of these bodies. 



The orange substance, being very soluble in water, was easily 

 separated from the volcanic sand with which it was mixed, and 

 which amounted to from 8 to 10 per cent. The soluble part I 

 found to be composed of metallic, earthy, and alkaline muriates, 

 the proportions of which varied in different portions of the 

 mass. One portion contained^ 



99.47 



It is obviously a mixture, and not a definite compound, but 

 I am not aware that any such mixture has hitherto been obser- 

 ved. I should conjecture that it has been formed by the passing 

 of gaseous chlorine over metallic matter at a red-heat, from 

 the resemblance it bears to the sublimate obtained when chlorine 

 is passed over red-hot iron-filings in a tube. The colour is ren- 

 dered paler, however, by the presence of the colourless mu- 

 riates. 



We may expect much light to be thrown on the theory of 

 volcanic eruptions, if those who have the opportunity will fol- 

 low the example of Mr Allan, and collect the products of such 

 eruptions, before they are decomposed by the atmosphere, or 

 washed away by the rains. 



