[ 327 ] 



XXXII. 



NOTES FEOM THE PHYSICAL LABOKATOKY OF THE EOYAL 

 COLLEGE OF SCIENCE : ON THE DETEKMINATION OF 

 THE ABSOLUTE EXPANSION AND THE DENSITIES OF 

 LIQUIDS. By Professor W. F. BAKEETT, Eoyal College of 

 Science, Dublin. 



[Read January 9, 1889.] 



The well-known arrangement of Dulong and Petit for determining 

 the absolute or true expansion of liquids (wherein the dilatation 

 of the vessel holding the liquid is eliminated) consists essentially 

 of a U tube filled with the liquid under examination, the two limbs 

 of the U being maintained at a constant difference of temperature. 

 This method depends on the hydrostatical law that the heights of 

 any two or more columns of liquid in communication, when in 

 equilibrium, are inversely as the relative densities of the liquids. 

 As the height is independent of the sectional area of the vessels 

 holding the liquids, the expansion and consequent increased capa- 

 city of the glass on the warm side does not affect the result. To 

 prevent the mixing of the warm and cold liquid column the com- 

 municating tube is made of very small bore. There is, however, 

 a slow interchange of liquid besides other experimental difficulties 

 connected with this arrangement. Regnault greatly improved the 

 apparatus by using an air space in the communicating tube between 

 the two liquid columns, whereby the fluid pressure was transmitted, 

 but mixing prevented. 



The arrangement I have devised is in principle the same as 

 Regnault's, but it has the advantage of being very simple and 

 inexpensive, and as it has been for some time in use in my labora- 

 tory, with most satisfactory results, it may be useful to make it 

 more generally known. 



Two glass U tubes are connected by a cross piece, furnished with 



