ELECTRICAL RESISTANCE UNDER PRESSURE. 91 



the relation between temperature and resistance was linear between 0° 

 and 100°. Here we have the unusual case of the presumably purer 

 substance with a smaller temperature coefficient. Recent observa- 

 tions of Holborn ^ make it likely, however, that the behavior of 

 aluminum is similar, and it is not surprising if the behavior of magne- 

 sium and aluminum should be the same. The average pressure 

 coefficients of the first sample between and 12000 kg. were —0.05446, 

 441, and 436 at 0°, 50°, and 100° respectively. This is seen to be 

 about 10% higher than that of the second sample. The two samples 

 agree, however, in having a coefficient which decreases with rising 

 temperature. The maximum departures from linearity of the im- 

 purer sample were also nearly the same as for the purer,* the maxi- 

 mum deviations being 0.0023, 0.0026, and 0.0029 at the three tem- 

 peratures respectively. The deviations from linearity of the first 

 sample were symmetrical with respect to pressure, in each case the 

 relation between pressure and resistance being representable by a 

 second degree equation, and the pressure of maximum deviation 

 being at 6000 kg. The first sample differs from the second in not 

 showing a closer approach to linearity at 50° than at the other two 

 temperatures. 



These new and more accurate results differ considerably from the 

 rough value given for the first sample; it was stated in the previous 

 paper that the average coefficient for the first sample was probably 

 - O.O555. 



Calcium. This material was obtained from the Research Labora- 

 tory of the General Electric Co. through the kindness of Dr. Langmuir. 

 An analysis by Mr. N. S. Drake by the method of differences showed 

 not more than 0.1% total impurity, the error of the measurements 

 being 0.1%. Qualitative analysis showed no detectible impurity 

 except a trace of iron, too small to determine quantitatively. 



The calcium was furnished in the form of a solid ingot about 1 inch 

 in diameter. Pieces of the appropriate size were cut from it with a 

 hack saw, and were formed into wire 0.013 inches in diameter by 

 extrusion through a steel die. An extrusion pressure of the order of 

 10000 kg per cm^ was required. A phenomenon shown to a more or 

 less pronounced degree by all metals during extrusion was particularly 

 prominent with calcium. If the extrusion pressure is pushed too high, 

 or sometimes for no apparent reason, the wire will suddenly break and 

 the metal spit out of the extrusion block in long gulps or small pieces. 

 This spitting forth reaches explosi-\'e violence in the case of calcium, 

 and the fine dust into which the issuing wire breaks takes fire spon- 



