Notices respecthig New Books. 57 



enclosure of which we desire to know the temperature, this being 

 indispensable ; but it presents a great inconvenience when the ap- 

 paratus is intended to measure elevated temperatures, for then the 

 greater portion of the air is contained in the calibrated tube, and 

 only a very small portion remains in the reservoir properly so called. 

 The result of this is that the portion which will pass out by a new 

 increment of temperature will be very small, and will be difficult to 

 measure with sufficient precision in the calibrated tube. The appa- 

 ratus thus becomes but very slightly sensible at elevated tempera- 

 tures ; M. Regnault has therefore rejected this arrangement of the 

 air-thermometer. 



In the second method the gas is maintained constantly in the 

 same volume, and the elastic forces which it presents under different 

 circumstances are the points determined. In this way, by knowing 

 the variations produced in the elastic forces, we may calculate, by 

 Mariotte's law, the dilatations which the gas would have experienced 

 if the pressure had been kept constant. The apparatus founded on 

 this method, besides being more easily manageable and of greater 

 precision, has the advantage of presenting as much sensibility at 

 high temperatures as at low. But in the employment of this me- 

 thod two important questions present themselves for solution. The 

 first is, to know if air-thermometers, Jilled with air at very different 

 densities, are comparable with each other. The second, if gas-ther- 

 mometers, jilled with gases of different nature, proceed in agreement 

 with each other when they have been regulated for the fixed points of 

 0° and 100°. 



The apparatus which M. Regnault has made use of to solve these 

 two questions consists of two gas-thermometers, each composed of Jl 

 glass balloon of from 700 to 800 cubic centimetres capacity, ter- 

 minating in a curved capillary tube, which abuts, end to end, against 

 another capillary tube communicating with the manometric appara- 

 tus. The union of the capillary tubes is effected by a brass tubulure 

 bearing a rectangular appendix, into which is cemented a capillary 

 tube, serving to put the apparatus in communication with an air- 

 pump, by means of which the interior of the apparatus may be dried, 

 and the different gases to be operated on be introduced. The bal- 

 loons themselves are plunged into a heating apparatus filled with oil, 

 which is constantly agitated in order to maintain an equal tempera- 

 ture throughout the bath. We pass over the details of the experi- 

 ments, and the enumeration of the precautions taken to dry the in- 

 terior of the apparatus properly, to maintain the temperature as 

 nearly stationary as possible, and to determine satisfactorily the 

 levels of the mercury in the manometers. The experiments were 

 conducted in the same manner, whether it was wished to compare 

 the course of an air-therniometer with that of a thermometer filled 

 with some other gas, or to compare the course of an air-thermome- 

 ter charged with air having an initial elastic force of about 760 

 milliras. at 0°, with that of a similar thermometer filled with air of 

 a less or greater density. 



The tables contain the results of the experiments made on the 



