On Tests for Arsenic. 431 



Is caloric itself, rendered visible by the sudden approxima- 

 tion of its parts in a small space, where it rises to a tem- 

 perature that is increased in the oxygen so as to kindle 

 ihe spunk*. 



I am equally induced to believe, since the air (and it i$ 

 the same with all gases) is decomposed by rapid com- 



Eression, that the luminous meteors frequently perceived in 

 urricanes are not always the effects of electricity. I have 

 observed several times, on these occasions, that Saussure's 

 atmospheric electrometer affords no signs of any. I will 

 mention a particular instance, as it occasioned me no less 

 surprise than damage. 



In the beginning of the year 1803, being at my country 

 peat, toward evening a violent wind arose, which continued 

 increasing for two hours to such a de.jree, as to blow down 

 al)out sixty trees of prodigious size and height in an orna- 

 mental plantation. It tiirevv thetn one upon another in a 

 row, and some of them were broken f)ff. Those that were 

 torn up by the roots brought \.\\t the earth with them to the 

 distance of fifteen feet. The clouds flew with extreme 

 rapidity, and I saw flashes of light from them. I raised 

 my electrometer, armed with its^conductor two feet long, 

 but the bails still continued in contact. 



I^XXVI. Mr, Hume's I^emarks on Dr. Roget's Reply, 



To Mr. Tilloch. 



Sir, Ur. RopET having confessed that I did make use of 

 the words "or anv other alkali," the question at isstie 



between us becomes more simple; it is merely this, 



Whence are the distinctions between sulphate of copper and 

 nitrate of silver, which prove that one of these salts ii 

 /o^/% different from the other, when applied for the detec- 

 tion of arsenic ? 



That there is a total difference, especially when ammonia 

 is employed, has been asserted by Dr. Roget ; — that there 

 is no such dissimilarity in the application of these two 

 tests, to support such an opinion, 1 shall now endeavour to 

 prove. 



•It sometimes happens, tliat the spunk is turned hiack without kiudlinc 

 In this case, as well ai when it is kinUlefJ. if we draw back the piston in 

 the tube, a dense vapour, that maybe s.nclt. issues out, which is not of the 

 •amc nature as the lonjur. -Jliat shows itself before the inllammation 5 thi» 

 always succeeds it. Th:.! h the principle of the inHammation : this is a pro. 

 0utt furawlicd by tht cwubuitioo yl the spunk, of which it h»» the iiuell. 



Let 



