190 ROYAL SOCIETY OF CANADA 



Again, with Eegnault's apparatus, the arrangement for measuring 

 xhe initial temperature of the gas was somewhat defective. On account 

 of the very rapid fall of temperature along the tube through which the 

 gas passed from the oil bath to the calorimeter it was scarcely penniss- 

 able to assume that it entered the calorimeter at the temperature of the 

 oil bath. 



II. — The Bunsen Ice Calorimeter. 



A modified form of the calorimeter used by Eegnault for the deter- 

 mination of the specific heat of gases is also used to find the specific heat 

 of liquids or solids. Another form of calorimeter which has been used 

 for liquids and solids but not for gases is the Bunsen Ice Calorimeter. 

 The construction of this type of calorimeter is shewn diagrammatically 

 in Fig. 1. A glass test tube B is sealed into the upper end of a larger 

 cylindrical glass vessel A. The lower end of A is joined to a U tube AC 

 which carries a cup C at its upper end. A hollow stopper furnishel 

 with a three way tap D prolonged into a graduated capillary tube E is 

 fitted into this end. By means of this tap communication can be madf 

 from the cup D, which the stopper carries, either to C or to E or from C 

 to E. 



In setting up the instrument for use the upper part of A was filled 

 with pure distilled water free from air, and lower part of A, the tube C 

 and part of the capillary with pure boiled mercury. By means of the 

 tap D the end of the mercury thread could be moved to any selected pos- 

 ition. 



A part of the water in A was frozen and formed into a cap of ice 

 around the immersed part of B as indicated in the diagram. In mak- 

 ing all measurements. A, B, and the lower part of were kept surround- 

 ed by ice in order to maintain the apparatus at zero temperature. 



The measurement of heat by this calorimeter depends on the fact 

 that water changes its volume on solidifying. If a quantity of heat is 

 communicated through B to the water in A a certain quantity of ice will 

 be melted. This will cause a diminution in the volume of the contents 

 of A, and a consequent receding of the mercury thread in E. From the 

 known values of the latent heat and the specific gravity of ice, the 

 amount of heat communicated may be determined for any change of 

 volume produced. 



An investigation with this type of calorimeter shewed that its deli- 

 cacy was such that an addition of one calorie made a change of about 

 1 1-3 millimeter divisions in the position of the end of the capillary 

 thread. With the apparatus used by Eegnault, on the other hand, an 

 addition of about 600 calories was required to make a difference of on*» 



