Ox THE MECHANICAL EQUIVALENT OF HEAT 429 



the method shown in Fig. 9. At the edge of the wheel, which was of 

 the exact diameter of the calorimeter, two screws were attached, from 

 which wires descended to a single screw in the edge of the calorimeter. 

 Through the wheel, a screw armed with a vulcanite point pressed upon 

 the calorimeter, and held it firmly. Three of these arrangements, at 

 distances of 120, were used. To centre the calorimeter, a piece of 

 vulcanite at the centre was used. By this method of suspension very 

 little heat could escape, and the amount could he allowed for hy the 

 radiation experiments. 



The Torsion System 



The torsion wire was of such strength that one millimeter on the 

 scale at the edge of the wheel signified 11-8 grammes, or ahout y^ of 

 the weights op generally used. There were stops on the wheel, so 

 that it could not move through more than a small angle. The weights 

 were suspended by very flexible silk tapes, 6 mm. or 8 mm. broad and 

 0-3 mm. thick. They varied from 4-5 k. to 8-5 k. taken together. The 

 shaft, ab, was of uniform size throughout, so that the wire c suspended 

 the whole system, and no weight rested on the bearings. 



The pulleys, m, n, Fig. 6, were very exactly turned and balanced, and 

 the whole suspended system was so free as to vibrate for a considerable 

 time. However, as will be shown hereafter, its freedom is of little 

 consequence. 



The Water Jacket 



Around the calorimeter, a water jacket, t u, was placed, so that the 

 radiation should be perfectly definite. During the preliminary experi- 

 ments a simple tin jacket was used, whose temperature was determined 

 by two thermometers, one above and the other below, inserted in tubes 

 attached to the jacket. 



The Driving Gear 



The cog-wheels, g, h, were made by Messrs. Brown and Sharpe, of 

 Providence, and were so well cut that the motion transmitted to the 

 calorimeter must have been very uniform. 



The Chronograph 



The cylinder of the chronograph was turned by a screw on the shaft 

 ef, and received one revolution for 102 of the paddles; 155 revolutions 

 of the cylinder, or 15,810 of the paddles, could be recorded, though, 



