Refe arches on Alumint. j$$ 



Thefe experiments were each made on 50 grains only, and 

 with a balance which 1 cannot warrant to ha\e been free 

 from an error lefs than a quarter of a grain. They are far, 

 therefore, from that degree of precifion which is necelfary to 

 determine the law of the progreifion in queftion. 



Thefe refults, in regard to the uniformity of their progrefs, 

 do not agree with Wedgewood's pyrometer: but I muff ob- 

 ferve, that this inftrument is too uncertain to be employed 

 for determining the law of deficcations. 



C. Necker, to whom I am indebted for the ufe of that 

 which I employed, expofed to the fame heat, and for the 

 fame time, eight pyrometric cylinders in a crucible of pla- 

 tina. He faw with me that they indicated 137—126 — 1 11 — 

 107—98 — 96 — 93 — 90 degrees of their fcale. Some of them 

 were vitrified ; others remained in the ftate of porcelain, or 

 changed their form more or lefs ; others experienced none of 

 thefe accidents. Thefe refults arife from the inequality of 

 the mixture of which the pafte of the cylinders is formed. 

 It can hardly be fuppofed that two artifts^ though furnifhed 

 with the fame formula, can be able to make limilar cylin- 

 ders, becaufe of the different degrees of the pulverization of 

 the filex, which will never be the fame, and will always have 

 a great influence on the refult. C. Gazeran obtained this 

 equality, which he even furpaffed, only by laborious and 

 multiplied trials, though founded on an analysis of Wedge- 

 wood's cylinders*. 



It is here feen, that if the property which alumine pofTefTes 

 at lovv temperatures, of lofing only a determinate quantity of 

 water at a conftant degree of heat, continues unimpaired at 

 a more elevated heat, it is poflible that, by employing for 



a doubt in regard to low temperatures ; but I think it in a great part er- 

 roneous, in regard to the fhrinking of pyrometric clay at a temperature 

 above the 29th degree of that inftrument. The experiment on which I 

 found this idea is as follows: — 1 weighed, with a balance feufible to the 

 eighth of a grain, 'he pyrometric piece of the lalt experiment. Before put 

 into the fire, it weighed 32-- grains. I took it out at the 29th degree, and 

 it had then loft z\ grains. After that term to the 170th of the fame fcale 

 it did not fenfibly change its weight, and yet during that interval it had 

 loft more than the fourth of its volume. This fh inking, therefore, does 

 not arife merely from the expulfion of the water, but rather, and aimoft 

 exclusively, from the new combinations or modifications which the earthy- 

 elements of the argjl experience at a high temperature. 



At the 170th degree, the platina crucible containing the alumine gave, 

 at its exterior furface, proofs of very grevit emollition, and aimoft of fufion. 

 This furface, in fome of its parts, exhibited a confufed cryftallizatiov fimi- 

 lar to that exhibited by Carrara marble when poiifhed. The platina co- 

 vering, though very thick, funk down, and was flightly fufed in fome 

 places. 



■ Philofophical Magazine, Vol. IX. p. 155. 



each 



