446 Inquiries info the Laws of Dilatation of Solids, Liquids, 



that the excess of the dilatation of the copper over that of the 

 glass remained Tcry nearly in proportion to the indications of > 

 the mercurial thernionicter to the temperatnre of 300" and up- 

 wards. This result, very different from that which. we had ob- 

 famed with the measurements of coj^per and platina, as it ap- 

 j)eared to us, could only be accounted for bv supposing the glass 

 to possess a law of dilatation still more ra})id than that of the ■ 

 metals. 



Before us, M. De Luc thoufdit he perceived in this substance a 

 similar property, although his experiments had not gone beyond 

 100^. His apparatus was composed of two vertical measuring 

 rods, one of glass and the other of i)rass, fastened together in- j 

 varial)ly at their lower extremities. He gave to these measuriiig i 

 rods lengths which are in the inverse ratio to their respective j 

 dilatability : the largest, which was the glass one, is kept fastened j 

 by its upper part ; the highest extremity of the copper rod is ? 

 entirely free ; and it results from the relation of the dimensions |! 

 of the two measuring rods, that this extremity cannot undergo 

 any change of situation at whatever temperature the measuring 

 rods are exposed, if the dilatabilities of the 'glass and of the 

 copper vary in proportion. Now De Luc had observed that 

 when he had established the compensation for a certain change 

 of ten)perature, there was no longer any room for greater or 

 smaller variations. 



But we must observe, that the measuring rods were in a ver- , 

 tical situation, and plunged into a vessel full of water, of which 

 the temperature was gradually raised. Although care was taken 

 to shake the liquid, it is probable that the inferior strata were- 

 always colder than the upper; and as the copper rod occupied i 

 the lower half, while that of glass passed through it in its whole- 1 

 height, we may ascribe the increasing dilatability which the glass j 

 has presented, to the copper rod being always found at a lower I 

 temperature than the other, and the differenc)e must have in- I 

 creased with the heating. We might besides remove this doubt, Ij 

 by repeating the same observation, the measuring rods being ii 

 placed horizontally, or rather by reversing the position of the 

 apparatus. 



M. De Luc, who does not seem to suppose that his apparatus i| 

 has the inconvenience we have mentioned, employed no means Ij 

 of verification ; so that we might fairly entertain some well founded I 

 doubts as to the result of his experiments. Witli the view of 

 verifying the consequences which we may deduce from the ob- 

 servations of this experimentalist and our own, we have endea- 

 voured to establish this same comparison by a more simple me- 

 thod. 



The 



