144 M. V. Regnault on the Expansion of Gases. 



the tube surrounding it. Owing to this second expansion and 

 the increase of vis viva, there is a great diminution in the tempe- 

 rature of the air before and around the thermometer, of the ex- 

 panded air; and that explains the excess of cold indicated by this 

 thermometer. 



In my experiments the first expansion was greater than the 

 second ; but this might have been easily reversed by increasing 

 the section of the capillary orifice. 



The experiments I have made with thermoelectric elements 

 have led to the same results ; they have, moreover, enabled me to 

 follow far more easily the variations in temperature which the 

 gaseous current undergoes. I shall only mention a few. 



The expansion takes place in a capillary glass tube 4 centims. 

 in length, terminated by a glass tube 30 millims. in internal 

 diameter, in which is arranged one of the elements, which can 

 thus be brought more or less near the capillary orifice ; and the 

 soldering may even be placed in this orifice. 



1. The thermoelectrical solderings exhibit differences of tem- 

 perature which are gradually smaller in proportion as one of the 

 solderings is removed from the orifice of the capillary tube. 



2. The greatest difference of temperature is observed when 

 this soldering is quite in the interior of the capillary tube, the 

 section of which it considerably diminishes. It is evident, more- 

 over, that, if the section were diminished beyond a certain limit, 

 the difference of temperature would, on the contrary, be dimi- 

 nished ; for if the orifice were entirely closed, the two solderings 

 would be at the same temperature. 



To realize still more closely the conditions in which Messrs. 

 Thomson and Joule observed a great increase of temperature, I 

 arranged the following experiment. 



In the tubulure of the large worm I fit a thick disk of 

 gutta percha, perforated in the centre by a small aperture sur- 

 mounted by a hollow cone. The thermoelectric element is sealed 

 in the axis of a gutta-percha cylinder terminated by a projecting 

 cone, which exactly fits in the hollow cone of the thick disk. The 

 soldering of the element is just level with the apex of the cone. 

 It is needless to say that the disk of gutta percha is supported by 

 a metal plate with a large aperture, which prevents it from being 

 deformed and even torn by the pressure of the compressed air. 



We here again see the lowering of temperature rapidly dimi- 

 nish as the soldering is removed from the aperture of the disk. 

 It rapidly increases, on the contrary, when the cone which enve- 

 lopes the soldering is pressed into its position. 



The greatest degree of cooling is obtained with a certain ap- 

 proximation of the two cones. When they are brought nearer, 

 the cooling diminishes ; it also diminishes when they are moved 



