121 
heat we wish to determine, and have the vessel surrounded 
by the usual non-conducting material, we can have a large 
rise of temperature in a considerable space of matter. Of 
course at the present time, when the specific heats of sub- 
stances such as silver, copper, and iron, have been very 
accurately determined, we do not need to have a water or 
absolute calorimeter, but only an instrument which will 
accurately register the rise of temperature and the value of 
whose graduations may be determined by standard metals ; 
so that although we do not know the absolute calorimetric 
value of the heat communicated to the air and glass of such 
an instrument, yet if its indications are proportional to the 
amount of heat communicated, we have all that is required 
to form a relative calorimeter. Then as to the construction 
of such an instrument. It was found that the use of ground 
joints in the instrument might be entirely avoided, as a red 
india-rubber stopper can be adjusted with nearly as much 
accuracy as a ground joint. 
The apparatus shown below as Fig. I. is the first form of 
