and clastic Fluids, and 07i the Measiireme7it of Temperatures. 443 



and machines, is very important in a great number of circum- 

 stances. The utility of these determinations has been at all times 

 felt ; and thev have employed methods more and more exact, in 

 proportion as the want of precision was ielt in physical researches. 

 We mav cite as most deserving of confidence the results of 

 Sm.eaton, Ramsden, Lavoisier, and Laplace. 



In the labours wliicii have been undertaken on this subject, 

 we have principally endeavoured to ascertain the absolute dila- 

 |;atious from 0° to 100 % and this is in fact sufficiejit for commoa 

 use. Messrs. Laplace and Lavoisier have besides ascertained, 

 that between the freezing and boiling point the dilatation of the 

 metals was exactly iu proportion to the indication of the mer- 

 curial thermometer. Borda also made the same comparison, 

 but in an extent of thermometrical scale too inconsiderable to 

 warrant any induction relative to the laws of dilatation. Never- 

 thelesis the oljject of his labour was fulfilled, since he proposed 

 merely to ascertain the variations of length wliich were produced 

 by the changes of temperature of the air in the instruments 

 which were to be employed in the measurement of the meri- 

 dian . 



The greatest dilHcultv experienced by all those who have been 

 occupied with inquiries into the dilatation of solids, j)roceeds 

 from the necessity in which we are of rendering one part of the 

 system absolutely fixed, v.diich implies much complication in 

 the a)jparatus. When we confine ourselves, as Borda did, to 

 measure the difference of dilatation of two metals, we are merely 

 obliged to render the instruments invariable with respect to each 

 other, a coiulition which it is much more easy to perform than 

 the first. 



As we only seek to compare between each other the laws of 

 dilatation of various bodies, it appeared to us that we should 

 equally well attain this end, by substituting the difference of the 

 dilatations for the absolute dilatations. 



It is chici'iv from this consideration that we determined to em- 

 ploy the following ajjimratus : 



This a|)paratr.s is composed of two rods, one of platina, the 

 other of red copper, twelve decimetres lojig, twenty-five milli- 

 metres broad, and four millimetres thick: they are connected in 

 an invariable manner by one of their extremities, by means of 

 an iron Iruucrse U|)on which they arc firmly screwed. To the 

 other extremity are ada))led brass rods, which rise at first verti- 

 cally and afterwards, bend horizontally. One of these rods has 

 a scale graduate(l into fifths of a millimetre, and the otiicr is 

 furnished with an index which marks the twentieths of the di- 

 vision of the scales, which admits of our appreciating the cen- 



ticmes 



