96 BKIDGMAN. 



liminary attempt was made to avoid the error hy using petroleum 

 ether as the medium, which is very much more fluid than the Nujol. 

 This was unsuccessful because of chemical action between the petrol- 

 eum ether and the calcium. Readings of the expansion of the cal- 

 cium and Nujol were made at 0°, 25°, and 50°. At 50° a very few 

 exceedingly minute bubbles appeared, probably the beginning of 

 chemical action. Accordingly the readings at 50° were discarded. 

 The auxiliary data needed in the computation were obtained from the 

 dilatations when the dilatometer was entirely filled with Nujol and 

 with mercury respectively. The volumes of the dilatometer and the 

 capillary were obtained by weighing the mercury required to fill them. 

 The bore of the capillary was calibrated for uniformity by conventional 

 methods. The density of the calcium was obtained by weighing it in 

 air and under Nujol. The densities given are corrected for vacuum. 

 The following values were obtained for calcium : 



Density at 21° 1.5563 



Coefficient of volume expansion, 0° to 21°, 0.000075. 



The values for Nujol were obtained incidentally and are recorded. 



Density at 21° 0.8786 



Coefficient of volume expansion, 0° to 21°, 0.000717. 



Strontium. Particular interest attaches to this metal because of 

 the fact that it is underneath calcium in the periodic table, and calcium 

 is unusual in ha\'ing a positive coefficient. The material I owe to Dr. 

 G. E. Glascock, who very kindly placed at my disposal some of the 

 material whose preparation and properties he has described. ^^ The 

 metal was prepared by the electrolysis of the fused salts. It was in 

 the form of fused nodules of sometimes two or three cubic centimeters 

 volume, and had been kept since preparation under kerosene. Some 

 sort of action had taken place between the metal and the kerosene, 

 under which most of the kerosene disappeared, and the metal became 

 coated with a fine gray powder. The action of calcium on kerosene 

 I have found to be very similar. On scraping off the gray powder, 

 the coherent metal is found underneath. On cutting into the cleaned 

 nodules with a cold chisel, slag-like inclusions are sometimes found. 

 Lithium prepared by electrolysis shows the same appearance. By 

 careful selection it was possible to find pieces large enough free from 

 these inclusions. 



The measurements were made on the metal in the form of wire 

 approximately 0.020 inches in diameter, formed by extrusion from one 



