OF ARTS AND SCIENCES. 157 



In my apparatus one of the new features lias been the introduction 

 of the Joule calorimeter in the place of the friction cylinders of Hirn 

 or the cones of Puluj. At first sight the currents and whirlpools in 

 such a calorimeter might be supposed to have some effect ; but when 

 the motion is steady, it is readily seen that the torsion of the calorim- 

 eter is equal to that of the shaft, and hence the principle must apply. 



This change, together with the other new features in the experi- 

 ments and apparatus, has at once made the method one of extreme 

 accuracy, surpassing all others very many fold. 



2. General Description. 



The apparatus was situated in a small building, entirely separate 

 from the other University buildings, and where it was free from dis- 

 turbances. 



Fig. G gives a general view of the apparatus. To a movable 

 axis, a b, a calorimeter similar to Joule's is attached, and the whole is 

 suspended by a torsion wire, c. The shaft of the calorimeter comes 

 out from the bottom, and is attached to a shaft, e j\ which receives a 

 uniform motion from the engine by means of the bevel wheels g and h. 

 To the axis, a b, an accurately turned wheel, hi, was attached, and the 

 moment of the force tending to turn the calorimeter was measured by 

 the weights o and p, attached to silk tapes passing around the circum- 

 ference of this wheel in combination with the torsion of the suspend- 

 ing wire. To this axis was also attached a long arm, having two 

 sliding weights, q and r, by which the moment of inertia could be 

 varied or determined. 



The number of revolutions was determined by a chronograph, which 

 received motion by a screw on the shaft ef, and which made one 

 revolution for 102 of the shaft. On this chronograph was recorded 

 the transit of the mercury over the divisions of the thermometer. 



Around the calorimeter a water jacket, t it, made in halves, was 

 placed, so that the radiation could be estimated. A wooden box sur- 

 rounded the whole, to shield the observer from the calorimeter. 



The action of the apparatus is in general as follows. As the inner 

 paddles revolve, the water strikes against the outer paddles, and so 

 tends to turn the calorimeter. When this force is balanced by the 

 weights o ;j, the whole will be in equilibrium, which is rendered stable 

 by the torsion of the wire c d. Should any slight change take place 

 in the velocity, the calorimeter will revolve in one direction or the 

 other until the torsion brings it into equilibrium again. The amount 



