576 



BRIDGMAN. 



grooves of a core about 9 cm. long. The usual accidents were 

 encountered in making these cylinders; several broke during use 

 because of defective steel. 



It was my original intention to make several apertures in the 

 lower cylinder, so that measurements could be made on several 

 coils at once, but this scheme was given up because two such cylin- 

 ders, after much work had been put on them, proved defective. It 

 seems to be the part of experimental economy when much effort in 

 making the apparatus may go for nothing because of defective mate- 

 rial, to make the apparatus of as simple design as possible, even at 

 the expense of extra labor in obtaining the readings. Positive results 

 were obtained with one of these preliminary cylinders, however, 



Figure 1. The cylinder to contain the wire under measurement. At the 

 top, connection is made to the pressure pump; at the bottom, the insulating 

 pkig with the wire is inserted. 



in which two coils could be placed simultaneously. Measurements 

 were made on two coils of iron, and at another time on two coils 

 of copper. The agreement between the members of these two 

 pairs of coils was perfect within the sensitiveness of the measure- 

 ments, which for these coils was about 1/3000 of the change of 

 resistance. This is gratifying because it shows that the effect of 

 pressure on resistance may be reproduced with not more error 

 than that in the absolute measurement of pressure. 



The resistance was measured by a null method on a Carey Foster 

 bridge. The galvanometer circuit was permanently closed to avoid 

 thermo-electric effects, and the battery circuit was closed only when 

 making measurements. The current was, however, kept so low that 

 the battery circuit could be kept permanently closed with no change 



