Chemical Sciente. 4TO 



the gas; for, as it enters the vessel from the parts* in which it was 

 previously confined, at a pressure of thirty atmospheres, it suddenly 

 expands, has its capacity for heat increased, falls in temperature, 

 and consequently cools that part of the vessel with which it first 

 comes in contact; but the part which has thus taken heat from the 

 vessel being thnist forward to the further extremity of the cylinder 

 by the successive portions which enter, is there compressed by 

 them, has its capacity diminished, and now gives out that heat, op 

 a part of it, which it had the moment before absorbed ; this it com- 

 municates to the metal of that part of the gas-vessel in which it is 

 so compressed, and raises its temperature. Thus the heat of tem- 

 perature is actually taken up by the gas from one end of the 

 cylinder, and conveyed to the other, occasioning the difference of 

 temperature observed. The effect is best observed when, as before 

 stated, the gas, at a pressure of thirty atmospheres, is suddenly let 

 into the vessels: the capacity of the parts is such, that the pressure 

 usually sinks to about ten atmospheres. 



6. On the Change in the Zero of Thermometers.* — M. Arago 

 remarks, relative to the gradual change which takes place in the 

 indications of the thermometer, in consequence of the elevation of 

 the mercury, that it throws much doubt over numerous observations 

 made with that instrument, and imperatively requires some process 

 by which it can be prevented. The process he recommends is one 

 already known to artists, and practised by them with alcoholic 

 thermometers, and consists in leaving air in the upper part of the 

 tube sufficiently compressed to prevent the disengagement of that 

 contained in the liquid, and the precaution he considers as not less 

 necessary, whether the metal really contains the air in solution, or 

 whether it be considered as interposed only between it and the 

 glass. 



Deluc has established, by numerous experiments, "that when 

 from any cause the particles of air disseminated in a liquid gather 

 together, they acquire a greater power to separate the particles of 

 the liquid ; and that this effect may be produced long before the 

 small collections of air become visible, and even before they are 

 capable of making their way through the fluid.'* The excess of 

 force spoken of by Deluc, is manifested by an augn>entation of 

 volume, and when the fluid is contained in the tube of a thermo- 

 hieter, by the ascent of the zero ; the effect is quick and considerable 

 in thermometers, containing alcohol, water or oil, and should be 

 smaller and less rapid in one containing mercury, though ultimately 

 an analogous result will be obtained. 



An experiment made by M. Arago is then described, in which 

 the effect of inclosing air in the thermometer is accurately ascer- 

 tained. Two similai- thermometers were constructed at the same 



• Quarterly Journal of Science, xiv. 411 j xv* 160, 868, 3TI. 



2 12 



