444 Inquiries into the Latos of Dilatation of Solids, Liquids, 



tiemes of the millimetre or the 120,000dth parts of the length 

 of the measuring rods. 



These rods rest on four copper rollers fastened to a bar of 

 iron. The whole machinery is placed in a vessel of red copper 

 fourteen decimetres long, fifteen centimetres deep, and ten cen- 

 timetres broad. We made use of fixed oil in the?e new experi- 

 ments as in those which were made upon the gases, and we had 

 recourse to the same means to reader the temperature siationary 

 during a time sufficient to permit the mea'^uring rods to acquire 

 an equilibrium of temperature with the liquid. There were also 

 two systems of metallic plates placed on each side of the mea- 

 suring rods, and which we might put in motion in the liquid 

 mass, so as to mix the different strata, and establish everywhere 

 a uniform temperature without any risk of the derangement of 

 the rods. Finally, the copper vessel was furnished with a lid 

 having four sockets, in which were thermometers which served 

 to indicate the differences of temperature which were expected 

 to take place in tlie various parts of the vessel. A thermometer 

 placed horizontally between the bars indicated the true tempera- 

 ture of the liquid. 



That we may avoid prolixity, we shall omit several trifling de- 

 tails, which concurred however to the exactitude of our observa- 

 tions. It is easy to see from the known dilatations of the platina 

 and copper, tliat a change of one degree in temperature ought 

 to vary the relative position of the index by about a hundredth 

 of a millimetre. Now it is impossible to be deceived with a part 

 of the index, however little a person may be versed in reading 

 off the divisions. This precision appeared to us sufficient in the 

 kind of researches which occupy us. Besides, in order to give 

 more sensibility to this instrnment, its dimensions must be aug- 

 mented, and then the difficulty of establishing a uniform tem- 

 perature throughout the whole extent of the measuring rods 

 would necessarily throw more uncertainty on the true dilata- 

 tions. 



As a point of departure, we have taken the state of the mea- 

 suring rods in an oil-bath, which remained several days in a 

 room the temperature of which did not sensibly vary. This 

 process appeared to us to be preferable to the employment of 

 ice, which does not present a temperature really fixed, unless we 

 can agitate it continually, particularly when the surrounding air 

 is from 15° to 20° above zero. 



The oil-bath was afterwards heated to the temperature of 100 

 degrees and upwards, with the same precautions as in the foregoing 

 experiments. The heat of the liquid mass and that of the fur- 

 nace being here very considerable, the maximum of temperature 



was 



