OF ARTS AND SCIENCES. 163 



The Engine. 



The driving power was a petroleum engine, which was very efficient 

 in driving the apparatus with uniformity. 



The Balance. 



For weighing the calorimeter, a balance capable of showing the 

 presence of less than -fa gramme with 15,000 grammes was used. 

 The weights, however, by Schickert, of Dresden, were accurate among 

 themselves to at least 5 nig for the larger weights, and in proportion for 

 the smaller. A more accurate balance would have been useless, as 

 will be seen further on. 



Adjustments. 



There are few adjustments, and they were principally made in the 

 construction. 



In the first place, the shafts a b and ef must be on line. Secondly, 

 the wheels m n must be so adjusted that their planes are vertical, and 

 that the tapes shall pass over them symmetrically, and that their edges 

 shall be in the plane of the wheel k I. 



Deviation from these adjustments only produced small error. 



(c.) Theory of the Experiment. 



1. Estimation of Work Done. 



The calorimeter is constantly receiving fTeat from the friction, and 

 is giving out heat by radiation and conduction. Now, at any given 

 instant of time, the temperature of the whole of the calorimeter is not 

 the same. Owing to the violent stirring, the water is undoubtedly at 

 a very uniform temperature throughout. But the solid parts of the 

 calorimeter cannot be so. The greatest difference of temperature is 

 evidently soon after the commencement of the operation. But after 

 some time the apparatus reaches a stationary state, in which, but for 

 the radiation, the rise of temperature at all points would be the same. 

 This steady state will be theoretically reached only after an infinite 

 time ; but as most of the metal is copper, and quite thin, and as the 

 whole capacity of the metal work is only about four per cent of the 

 total capacity, I have thought that one or two minutes was enough to 

 allow, though, if others do not think this time sufficient, they can 

 readily reject the first few observations of each series. "When there 

 is radiation, the stationary state will never be reached theoretically, 



