14 



Mr. Carl Barns on the 



coaxially with the steel piezometer PP, containing the oil 

 to be tested. Tubes of glass, g g, hold the rod s s in position. 

 Owing to the fact that above 200° glass conducts better than 

 most hydrocarbon oils, this insulating arrangement is not quite 



Fig. 4. — Arrangement for measuring the Pressure Increments of the 

 Electric Resistance of Liquid Insulators. 



satisfactory ; but by using thin- walled glass tubes filled with 

 the oil, the data so obtained are sufficient for the purposes of the 

 present text, since the rate of breakdown with temperature is 

 accentuated. It is seen that the measuring-current passes from 

 the piezometer P P, through the cylindrical layer of oil, into 

 the rod s s, and thence to the battery. The diameters of steel 

 core and piezometer-tube being '13 and '63 centim. respec- 

 tively, and the exposed part 9 centim. long, the specific 

 resistance of the layer of oil is easily computed (§ 19). 



6 c. Digression. — By filling the greater part of the closed 

 end of the inner tube (fig. 1 or 2) with the substance to be 

 examined, and the open end as far as e with mercury, the tube 

 is available for the study of melting-point and pressure. In 

 such a case the forward motion of the inner mercury meniscus 

 due to pressure proportionately decreases the resistance of the 

 arrangement by increasing the opposed surfaces of mercury 

 and sodium amalgam. Increased delicacy is secured by 

 enlarging the bulb at the end of the inner tube. At the 

 solidifying-point the sample usually undergoes a phenomenal 

 reduction of volume, corresponding, therefore, to a similarly 

 large reduction of resistance. This occurrence I propose to 

 use as the criterion of solidifying-point and of melting-point. 

 The apparatus is available above 100° under any pressure up 

 to several thousand atmospheres. 



The criterion specified is also available when a fine platinum 

 wire is stretched through contiguous coaxial columns of the 

 substance and of mercury in a vertical piezometer. 



7. Resistances were first measured by a bridge method ; 

 but owing to the fact that electromotive forces are involved 

 which may themselves vary with pressure, I lost faith in the 



