Oft the various Comhnatms of CatioH tulth Oxigin. 3 



TJie laft metallic oxide which I fubjefted to this procefs, was that of manganefe; about 

 half an ounce of the black ore of this femi-metal was expofed for an hour to a ftrong red 

 heat irt a crucible, it was then mixed with hot charcoal, and introduced into a coated glafs 

 retort ; heat being applied and gradually increafed, a quantity of gas, amounting in all to 

 38 ounces, was procured. The firft portion was chiefly carbonic acid ; the fecond a rtiix- 

 ture of this with inflammable gas, in nearly equal proportions ; and the laft almoft purely 

 inflammable. 



• From thefe experiments we may draw the following conclufions : ift. That all metallic 

 oxides capable of enduring a red heat, will, when mixed with charcoalj not only yield 

 carbonic acid, but alfo a very confiderable quantity of an inflammable gas. 2dly, That 

 thofe oxides which retain their oxigen moft obflinately, yield the greateft quantity of in- 

 flammable gas; and on the contrary, that thofe which part with it readily, afford the 

 greateft proportion of carbonic acid. And 3dly, That the carbonic acid gas was chiefly 

 difengaged at the beginning of the procefs, and the pureft and greateft quantity of inflam- 

 mable gas towards the conclufion of it. 



Having thus afcertained the relative proportions of thefe gafes, I next endeavoured to 

 inveftigate the properties of the inflammable gas, which from the peculiar circumftances 

 under which it was procured, I had little doubt muft be difi'erent froifl any of the known 

 hydrocarbonates. 



Accordingly I took a quantity of that obtained from the grey oxide of iron, and waflied 

 it repeatedly with lime water, until the whole of the carbonic acid was completely feparated. 

 Its fpecific gravity was then afcertained, and it was found to be but a very little lighter 

 than common air, that is to fay, in the proportion of about 22 to 23, a circumftance which 

 proved it at once to be different from the common hydrocarbonates, all of which are much 

 lighter than common air. 



J next endeavoured to eftimate the quantity of pure oxygen neceflary to faturate it, and 

 likewife the produdls of the combuftion. After feveral trials I found, that when four 

 meafures of the waflied gas, mixed with two of oxygen, were introduced into a ftrong 

 glafs jar over mercury, and fired by the eleftric fpark, they were reduced to 3I meafures, 

 and that when this refiduum was agitated with lime water, there only remained about ^ of 

 a meafure, which was found by the nitrous teft to be pure oxygen ; hence it would appear, 

 that eight meafures of this gas require 3 \ meafures of pure oxygen gas to faturate them, 

 and when combined produce nearly fix meafures of carbonic acid gas, with a little water. 

 Now if we fuppofe a meafure equal to a cubic inch, and that a cubic inch of the inflam- 

 mable gas, at the temperature of 55°, and under a mean preflTure of the atmofphere, weighs 

 •3 of a grain, a cubic inch of oxygen gas, .34 of a grain, and a cubic inch of carbonic acid 

 gas .47 of a grain; then the weight of eight meafures of the inflammable, with 3! of 

 oxygen gas, will be 3.6 grains very nearly ; but the weight of the carbonic acid gas gene- 

 rated, is only 2.8 grains; hence there remains a deficiency of .8 of a grain to be accounted 

 for, and which muft arife from the water produced. The chief circumftance, however, 



^ i which 



