PRESSURE ON RESISTANCE OF METALS. 585 



Detailed Data. 



The detailed data for individual substances follow. These are 

 arranged in order of mcltinij point, except for the anomalous substances 

 bismuth, antimony, and tellurium. 



Indium. A sample only one gram in amount was available from 

 ]\Ierck, without analysis. This metal is as soft or softer than lead. 

 It was extruded into a wire of O.OOG inch diameter in a die of special 

 construction. Indium oxidizes much less rapidly than lead; after 

 extrusion the surface of the wire is brightly polished and remains so for 

 at least several weeks when exposed to the air. It was wound loosely 

 on a bone core of the dimensions already given. Its actual resistance 

 at 0° was 11.7 ohms. Connections were made by soft soldering with a 

 miniature copper, using a fusible solder of melting point slightly 

 above 100°. There is some difficulty in making a successful soldered 

 connection because of the low melting point of the indium, which is 

 about 155°. It alloys very rapidly with any ordinary solder, forming 

 an alloy of much lower melting point than any of the constituents. 

 It must be caught by the solder with a single well directed touch. 



The melting point of the alloy at the soldered connections limited 

 the range over which measurements could be made. Successful runs 

 were made at 0°, 25°, and 50°, but at 75° the soldered connections 

 dropped off. Difficulty because of alloying also made it necessary to 

 omit the usual temperature seasoning; this in any event is not so 

 necessary for a low melting metal as for a higher one. No irregularity 

 to be ascribed to lack of seasoning was to be found. After the very 

 first application of pressure to 12000 kg. at 0° there was a permanent 

 change of zero of 1.3% of the total pressure effect. This may well 

 have been a viscosity effect from the petroleum ether, since indium is 

 the softest of all the substances tried and so is particularly susceptible. 

 At 25° the permanent change of zero after 12000 kg. was only 0.3% 

 of the pressure effect, and at 50° only 0.003%. The maximum devia- 

 tion of any reading from the smooth curves is 0.3% of the total effect, 

 and the average numerical deviation 0.13%. 



The smoothed results are collected in Table I and the experimental 

 points are shown in Figure 3. The average coefficient does not depart 

 more than 0.2% from linearity with temperature. The deviation from 

 linearity is not symmetrical about the mean pressure; it is so great 

 that the initial slope is from 25 to 33% greater than the mean and the 

 final slope from 12 to 16% less. 



