70C Miss Agnes M. ClerJce 



mutual action might hence undergo modifications, the intricate con- 

 sequences of which can only in part be divined. 



An investigation of the dielectric constants, or specific inductive 

 capacities, of frozen electrolytes was undertaken in 1897. It met 

 ■with numerous difficulties, but led to some important conclusions. 

 These may be summarised as follows. Such substances as ice and 

 alcohol are capable, at low temperatures, of acting as dielectrics, 

 notwithstanding that some of them possess, in the liquid state, rela- 

 tively high electrolytic conductivity. They have dielectric constants 

 of large value near their freezing-points, which are greatly reduced 

 by cooling down to — 200° C. At the absolute zero, these values are 

 probably equal, all alike being two or three times that of the dielectric 

 constant of a vacuum. Near this point, too, all electrolytes tend to 

 acquire infinite resistivity, or to become perfect non-conductors of 

 electricity. Finally, at very low temperatures, frozen electrolytes are 

 nearly perfect insulators, but they rapidly regain sensible conducting 

 power at temperatures far below their melting-points. 



Oxygen and air in the liquid state were inferred from their re- 

 markable insulating quality to be dielectrics ; and it was accordingly 

 desirable to ascertain their dielectric constants, in terms of that of a 

 vacuum taken as unity. They came out 1*493 and 1*495 respec- 

 tively. Between the magnetic susceptibility of gaseous and liquid 

 oxygen a significant difference was elicited. The ratio for equal 

 volumes proved to be 1594 to 1. The magnetic susceptibility, in other 

 words, of the gas was, for equal masses, nearly doubled by lique- 

 faction ; whence the inference was drawn that this property does not 

 simply appertain to " the molecule per se, but is a function of the 

 state of aggregation." Noteworthy, besides, was the verification for 

 liquid oxygen of Maxwell's law connecting magnetic permeability, 

 specific inductive capacity, and optical refractivity. Additional ex- 

 periments on liquid oxygen made in 1898, on a different principle 

 from that previously adopted, afforded a qualified confirmation to the 

 law that magnetic susceptibility varies directly as the density of the 

 paramagnetic body, and inversely as its absolute temperature. 



So far from showing any tendency to disintegrate into " cosmic 

 dust," matter grows continually more rigid with cooling. A metallic 

 rod will sustain, for the same extension, four or five times the weight 

 at — 182° that it would at 0° C. A coil of fusible metal wire, which 

 the tension of a single ounce would pull out straight at the ordinary 

 temperature, will support a couple of pounds and vibrate like a steel 

 spring after immersion in liquid air. The most definite means, how- 

 ever, of determining the changes in cohesive force produced by cold, 

 is to compare the breaking stresses of metals at moderate and very 

 low temperatures. The necessary experiments, it is true, involve the 

 expenditure of gallons of frigid and costly fluids ; they were, never- 

 theless, carried out satisfactorily at the Royal Institution in 1893. 

 They showed a large increase with cooling in the tenacity of all 

 common metals and alloys. Exceptions presented by castings of zinc, 



