VOL. LXIX.J PHILOSOPHICAL TKANSACTIONS. 537 



extracted from other substances besides metals, which may explode like that ex- 

 tracted from metals ; but I only say, that in those cases the inflammable air will also 

 sparkle more, and will be found less easy to be decomposed by water. There are 

 other substances that give the inflammable air in great quantity, and which cannot 

 be considered as animal or vegetable substances, but come rather near the nature 

 of metals; as, for instance, the spathose iron, from which I extract a good deal of 

 inflammable air by the action of fire only, applied to a matrass. But the metal 

 in this substance is not in its pure state, and it may be considered rather as a calx 

 of iron than true iron. Accordingly, this air can hardly sparkle at all; it 

 explodes more like the inflammable air of vegetable or animal bodies, than that 

 of metals, and it is easily decomposed in water. 



This property of inflammable air of metals, which I have discovered, throws 

 great light on the analysis of the decomposition of that air, which I have made 

 in 2 different ways. The first is to fire it together with common or dephlogisti- 

 cated air, in vessels filled with very pure quicksilver, and also in vessels filled with 

 distilled water. The 2d method is to decompose it by shaking it in pure distilled 

 water. In the first process a great number of experiments are required in order 

 to obtain a sensible residuum; besides, the igneous part is lost. The 2d method 

 requires an exceedingly long time, but it is the most complete; for which reason 

 I have used it for the decomposition of other kinds of air. 



XXV. On the Variation of the Temperature of Boiling ff'^ater. By Sir Geo. 

 Shuckburgh, F. R., and A. S., &c. p. 362. 



The heat of boiling water having for some years been used as one of the 

 terms for graduating the scale of thermometers; together with the particular 

 attention the society has lately given to this branch of inquiry; and I may add 

 the singular success with which this age and nation has introduced a mathematical 

 precision, hitherto unheard of, into the construction of philosophical instru- 

 ments; will render it unnecessary for me to say more in respect of the following 

 experiments, than simply to lay them before the r. s. 



That the heat of boiling water was variable, according to the pressure of the 

 atmosphere, seems to have been known to Fahrenheit as early as the year 1724.* 

 A few years after this, M. LeMonnierand Cassini made some decisive obser- 

 vations, to show that this quantity was very considerable. It was left however 

 for Mr. De Luc to make a much more complete series of experiments, which he 

 had described and reduced into system in his Recherches sur la Variation de la 

 Chaleur de I'Eau bouillante. It remained only that these should be verified. 

 Towards the end of 1775 I had an opportunity of repeating these observations, 



* Vide Phil. Trans., N" 385, where a curious project is proposed, of determining the weight of the 

 atmosphere by means of a thermometer alone, under the title of " Barometri novi descriptio." — Orig. 

 VOL. XIV. 3 Z 



