5 



1 e, 



Chemical Equilibrium of Solids. 13 



the outer compartment at m m (sodium amalgam) , with 

 petroleum at n n. The terminal wires are shown at bb 

 (cathode connecting with the piezometer) and at cc (anode, 

 insulated). 



() a. In the course of my work, however, I found it 

 essential to connect c c with the piezometer and to insulate 

 b b ; for by thus reversing the tube there is no tendency of 

 short-circuiting through oil and the outer tube. Indeed, so 

 important is this adjustment that I further modified the tub 

 fig. 1, in the way indicated in fig. 3. 



Fig. 3. — Improved Form of the Tube. 

 co LcL — >-* Jzot »■ 



->C 



Here the inner compartment is filled with mercury to only 

 about 7 centim. of the end. The remainder of the tube, 

 containing hydrocarbon oil, runs completely through the 

 piezometer into the barrel, with which the insulated platinum 

 terminal connects. To keep the mercury in place, the inner 

 tube is drawn down to a smaller diameter (see i, fig. 3) , into 

 which the mercury meniscus projects. Air is excluded by 

 boiling the mercury with the platinum wire in place. The 

 latter amalgamates externally, but careful handling is none 

 the less advisable. The outer tube, with its filling of sodium 

 amalgam, not only envelops the inner mercury-thread, but 

 extends in both directions as far as the cold parts of the sur- 

 rounding piezometer-tube. {Since c c connects with this, no 

 current will pass through the film of hot oil and the outer 

 glass tube. All currents between piezometer and barrel must 

 therefore pass through the walls of the inner glass tube — a 

 condition which is further insured by making NaHg/glass/Hg 

 the acting electromotive force. I have purposely described 

 these precautions at length ; and their importance is obvious 

 when it is remembered that the oil medium has a larger 

 pressure-coefficient than glass, and that the insulation of oil 

 sometimes breaks down in marked degree with rise of tem- 

 perature (§ 19). The effect of cold oil can be found pre- 

 liminarily, when the whole apparatus is cold. 



6 b. I have finally to show the apparatus for testing the 

 resistance and pressure-coefficient of the liquid insulators 

 at different temperatures. This is given in fig. 4 (diagram), 

 and consists essentially of a cylindrical steel rod s s, adjusted 



