EFFECT OF PRESSURE ON CONDUCTIVITY OF INIETALS. 89 



If the specimen were one of those with a small hole cast along the 

 axis the heating element was a single length of 0.0035 inch nichrome 

 wire, coated with enamel as the hairpin unit, silver soldered at one end 

 to a massive copper wire, and grounded at the other to the specimen 

 itself by soldering to a piece of fine german silver sheet, which was in 

 turn soldered to the specimen. By using an intermediate member of 

 fine german silver sheet the difficulty of attaching the fine wire to the 

 massive specimen at exactly the required point was obviated, and it 

 was easy to make the heating element the same length as the specimen 

 itself, at least within 0.002 inch, which was close enough. The speci- 

 men itself, which was grounded to the pressure cylinder, was used as 

 the return connection for the heating element. In order to eliminate 

 any danger of bad contact between the specimen and the cylinder, a 

 connection was soldered to one of the steel pillars and the ground 

 connection of the three terminal plug. 



The thermo-couple, as already mentioned, was constructed of copper- 

 constantan. Each couple was made of three pieces of wire, first a 

 length of copper, then one of constantan, and then one of copper, of 

 the same length as the first. This was bent into a hairpin form, and 

 the two legs of the pin slid through the two thermo-couple holes. The 

 dimensions were so chosen as to bring the two junctions copper- 

 constantan midway between the ends of the specimen. The two 

 copper ends of the couple were attached by solder to the two copper 

 thermo-couple terminals of the three terminal plug. It is thus seen 

 that the e.m.f. of the couple is determined by the temperature 

 difference of the two copper-constantan junctions. 



The wires of which the couple were made were 0.010 inch diameter. 

 It is essential that the junctions be absolutely smooth butt joints. It 

 is a little trick to successfully make these, but the following procedure 

 finally gave satisfactory results. The ends of the wires to be united 

 were first coated thinly with silver solder and borax, precisely as 

 ordinary objects are "tinned" for soft soldering, by dipping into a 

 drop of molten silver solder covered with a little borax. The drop of 

 silver solder was maintained molten at the desired temperature in a 

 miniature electric furnace made by winding nichrome wire around a 

 core (of course insulated from it) of j inch nichrome rod, in the end 

 of which was drilled a shallow depression to hold the drop of molten 

 solder. After coating the ends of the wires, they were brought into 

 exact axial alignment in a clamp with three adjustments. The wires 

 were held in pieces of quartz tubing drawn down to the proper size, 

 and projected perhaps | inch or less from the tube. The solder on 



