and elastic Fluids, and on the Measurement of Temperatures. 375 



lowed the same course in their dilatations, it will become extreme- 

 ly probable that these dilatations will indicate the true tempera- 

 tures : this is what might besides be verified by the correspond- 

 ing quantities of heat. By following this course, we shall have 

 the advantage of bringing into the determination of the true 

 thermometrical scale, all the precision which we may now give 

 in the measurement of dilatations; and these measurements them- 

 selves, independent of the important consequences which we 

 might deduce from them relative to the theory of heat, will pre- 

 sent data which will be useful under several circumstances. Such 

 are the considerations and the motives which determined us to 

 commence our labours by the comparison of the dilatations of 

 the gases and the solids with the mercurial thermometer in high 

 temperatures. 



Comparison of the Dilatation of the Gases and of the Progress 

 of the mercurial Thermometer. 



The apparatus which we u?ed was composed of a rectangular 

 tub of red copper, seven decimetres long, one decimetre broad, 

 and a decimetre deep. This tub has on one of its lateral faces 

 two c)|jenings, one of which serves to introduce, in a horizontal 

 situation, a mercurial thermometer, and the other holds the open 

 extremitv of a tube which is placed horizontally at the same 

 height with tl'.e thermometer. This tube is perfectly dry, and 

 contains air which is dry also. 



The tub rests on a furnace constructed in such a manner as to 

 heat all parts equally. It is filled with a fixed oil, which 

 may, as we knov,-, support a temperature of more than 300 de- 

 grees without boiling. The tube which contains the air is ter- 

 minated on the side of the aperture by a short tube of a very 

 small diameter, vvhich partly issues from the tub. The quan- 

 tity of air contained in the exterior portion of this tube, and 

 which does not participate in the heatitig of the rest, is not 

 worth noticing. We ascertained that it never exceeded a two- 

 thousandth part of the total mass ; and besides, we had the pre- 

 caution to heat it during every experiment, in order to reduce 

 the error which might result from it. 



The tub is covered by a lid pierced with several holes: some 

 are traversed by thermometers, which serve to indicate if the dif-r 

 ferent parts of the litiuid mass are at the same temperature; the 

 others have stalks or sticks terminated by vertical plates of cop- 

 per, which we can turn j wc thereby produce in the liquid a 

 brisk agitation, the object of which is to establish the uniformity 

 of temperature. 



The following -is the course pursued by us in every experii- 

 A a 4 ment. 



