2 Palmer — Pressure Coefficient of Mercury Resistance . 



Moreover it is much easier to prevent leakage of oil than 

 water and it has no detrimental effect on the steel parts with 

 which it comes in contact. An Amagat " manometre a pistons 

 libres"* was used to determine the pressures attained and gave 

 results within one-tenth per cent throughout the entire range. 

 The ratio of its pistons is such that one millimeter difference 

 in height of the mercury column supported by the larger cor- 

 responds to a difference of pressure of *64T atmospheres on the 

 smaller. It is supplied with an open glass manometer three 

 and one-half meters high and is therefore capable of indicating 

 pressures somewhat greater than two thousand atmospheres, but 

 above this limit a large element of uncertainty is introduced 

 by leakage of oil around the pistons. 



Commercial mercury was digested for about forty- eight 

 hours in a solution of sulphuric acid and bichromate of potash 

 in water and after being carefully washed, dried, and littered, 

 was distilled directly into the tubes in which its resistance was 

 measured. An ordinary glass thermometer tube, about 18 cm 

 long and 0*l mm bore, had 10 cm of 2 mm bore tubing welded to its 

 upper end in such a manner that a cavity, about l cm long and 

 4 mm in diameter, was formed between the two parts. This 

 cavity and the elongated bulb at the lower end of the fine 

 capillary had platinum electrodes melted through their walls 

 and, when filled, formed the terminals of the mercury thread, 

 under investigation. The open end of the large tube was 

 welded to a small glass mercury still connected, through a dry- 

 ing chamber, with a Geissler-Toepler air pump and the whole 

 apparatus exhausted until the pressure fell below one milli- 

 meter. When the inside walls had become perfectly dry, heat 

 was applied and mercury slowly distilled over and condensed 

 in the experimental tube. As soon as this became* full it was 

 emptied and the operation repeated until the mercury thread, 

 when examined with a magnifying glass, appeared perfectly 

 bright and uniform throughout its entire length. Air was 

 then admitted and the large capillary cut off about two centi- 

 meters from the point where it joined the still. After solder- 

 ing silk insulated copper wires to the electrodes the tube was 

 placed inside the steel piezometer and the upper wire con- 

 nected directly to it, while the lower one, after passing down 

 through a narrow glass tube, to insure good insulation, was 

 soldered to the inside of the compressor. Oil was forced up 

 into the piezometer and when its appearance at the top showed 

 that all air had been expelled the opening was closed by a 

 tinned steel screw. 



The piezometer was surrounded by a long brass cylinder, 



* Amagat, C. R., ciii, p. 429. 1886. Professor Tait has described a similar 

 apparatus in the '•Challenger Reports," 1873-16, Physics and Chemistry, vol. ii. 



