152 Lahy and Roberts: 



bearing of the stator, the rotor was balanced. Suitably illuniinated" 

 points situated on the plate in quevstion were observed with a- 

 naicrosoope, carried by a support free from vibration, and the 

 rotor was balanced until the amplitude (i.e., diameter of circum- 

 scribing circle) of the vibration of the plate was reduced to .001 

 cnii. 



The; field magnets have two j^oles, and there are two windings on 

 each pole. These windings can be connected in series so that the- 

 turns reinforce one another, and the niaxinuim flux for a given 

 current produced. They are so connected for the heavy experiment. 

 The windings can also be connected in series so as to oppose one- 

 another in their magnetising effect. They are so connected for the 

 light experiment. In this way the flux can be reduced to one-tenth 

 witliout changing the current in the windings and, therefore, with- 

 out changing tlie temi^erature of the rotor in the light and heavy 

 experiments. 



l^he Stator and Calorimeter. — The con.struction of the stator is 

 shown in Figs. 5, 6 and 7. Fig. 5 shows copper and iron cylinders. 

 The iron cylinder increases the magnetic flux, and supports the 

 copper cylinder. The channels on the external surface of the iron, 

 and the axial hole, carry the water in it« circulation througii the 

 apparatus. Ilie iron cylinder is attached t-o a glass tube (see Fig. 

 6) tilled with eider dow'n, in order to reduce loss of heat by con- 

 duction. The glass, tube in turn is attached to a steel shaft which 

 passes through a ball-bearing, and, at its upper end, is suspended 

 by a steel torsion wire. The wire supports the whole weight of the- 

 shaft and calorimeter. 



A thin sheet cylinder encloses the copper cylinder, and a Dewar 

 flask encloses the steel, copper, and iion cylinders, which to pre- 

 vent corrosion are all silver plated. 



The water enters the calorimeter (see Figs. 5, 6 and 7) through 

 a rubber tube, flows downwards ];etween the inside wall of the 

 Dewar flask and the thin steel cylinder, turns upwards and flows be- 

 tween the steel and the copper, then down between the copper and 

 iron, and finally out of the calorimeter through the axial hole in the 

 iron armature. The object of this somewhat elaborate circulation 

 is — (1) to bring the water into thorough contact wdtli the copper 

 and iron in which the heat is generated, (2) to break up stream 

 lines, and so ensure that the inlet and outlet water tenqieratures 

 determine the heat k)sses of the calorimeter. 



The ball bearing which maintains the shaft vertical was very 

 carefully made, and is used without oil lubrication, which w^as 

 found to increase the friction. Ihe angular amplitude of the stator 



