Mechanical Equivalent of Heat. 153 



when set in torsional oscillation decreases by 10 per cent, per 

 vibration, from which it can be shown that the friction is very 

 small. A mirror is attached to the shaft of the stator, and its 

 movement observed by lamp and scale. With the apparatus 

 arranged for a determination of the mechanical equivalent, the 

 stator oscillations are not critically damped; frictional resistance 

 arising from the viscosity of water is added, till the damping is 

 critical. 



Couple. — Two thin wires (see Fig. 1) pass from the circujiiference 

 of the wheel over two other wheels (see Fig. 2), and then to the 

 x\''eights. With the axis of the stator vertical (which it is to within 

 12' of angle) these wires should l;e parallel and horizontal, con- 

 ditions which are readily fulfilled. 



The design, construction and testing of the wheels shown in Fig*. 

 8 has requii'ed a great deal of attention. Ball, i-oller, and cone 

 bearings were tested, and found to po.ssess far too much friction to 

 be suitable for these wheels, and so a knife edge was used. While 

 this bearing is quite free from friction, and practical in use, it is 

 necessary to locate the position of the knife edge relative to the 

 centre of the wheel, a test which is not so eaisily made as might be 

 expected. 



Measurements. 



The relation 



J = ^J)nyim, - m,)/ -{ (M, - M,)(^, - 0,) |- 

 indicates what degree of accuracy is necessary in the various 

 quantities in the right hand niendier, if J is to be correct to 1 in 

 10,000. 



D, the diameter of the wheel, is 20 cm., and can be measured 

 to the necessary accuracy. The revolutions, n, are counted. The 

 acceleration of gravit;^, g, is accurately known for Melbourne. The 

 masses, ?»i and tHo gm., are readily found to the required preci- 

 sion. 



With O., — 0i = lO°C. it is necessary to determine this difference 

 to 1/1000'C, which is about the limit of accuracy obtainable with 

 a platinum, or, in fact, any thermometer. The water flows past 

 the platinum thermometers contained in Dewar flasks, as shown in 

 Fig. 2. 



The water, after passing through the calorimeter is collected in 

 a copper can. A two-way tap turns it into this can at the begin- 

 ning of the experiment, and at the same time starts the clirono- 

 graph record; at the end of the experiment turning the tap causes 

 the water to flow to a different vessel, and also stops the chronograph 

 record. The water collected is weighed. 



