ELECTRICAL RESISTANCE UNDER PRESSURE. 113 



put on this substance. The potentiometer method of measurement 

 was used. This substance cannot be soldered, so that it was neces- 

 sary to make connections with fine springs; sUpping of the springs 

 was perhaps accountable for some of the irregularities. The surface 

 layer has a very high resistance, which again introduced irregularities 

 at the spring contacts. 



Two runs were made on the impurer specimen, at 0° and 94°. The 

 irregularities of the second run were so great that the results were not 

 computed. Within the limits of error the relation between pressure 

 and resistance is linear at 0°, and the coefficient is — O.Oeo. 



The results on the second and purer sample were much more regular. 

 Readings were made at 0° and 95°. At both these temperatures the 

 relation is linear to 12000 kg. The best value for the pressure coeffi- 

 cient is - 0.06.398 at 0°, and - O.OeSOG at 95°. The coefficient is 

 seen to be very small; such small values have been found only for 

 certain of the high resistance alloys. The specific resistance of these 

 Zirconium filaments was also very high; 200 X 10"^ ohms/cm^ is the 

 value given me by the General Electric Co. 



Arsenic Considerable interest attaches to this element because 

 of its position in the periodic table above bismuth and antimony and 

 below black phosphorus, all of which are abnormal in behavior under 

 pressure. The arsenic used in this experiment was furnished by 

 Eimer and Amend. It had been distilled in vacuum, but was other- 

 wise of ordinary commercial ciuality, and I have no way of knowing 

 what the impurities might have been. I attempted to cast it in- a 

 mold of pyrex glass, supported on the outside with magnesia, and 

 enclosed in an iron pipe with caps tightly screwed on the ends. The 

 melting temperature of arsenic was high enough to melt the pyrex, 

 however, and the arsenic was found after the heating in the form of a 

 solid slug in the lower part of the magnesia powder. It may possibly 

 have come in contact to a sHght extent with the iron of the pipe while 

 in the molten condition. A slender rod about 1 mm. square in section 

 and 2 cm, long was worked out of the slug with a file and a hack saw 

 and by grinding. Grooves were filed on the ends, connections made 

 with spring clips, and measurements made with the potentiometer 

 in the regular way. I was surprised to find after I had completed 

 the measurements that Matthiesen ^^ had soldered connections to 

 arsenic, and I verified for myself that it is as easy to soft solder to 

 this metal as to antimony, for example. In fact the completely 

 metalhc character of the massive casting is a surprise contrasted with 

 the appearance of the sublimed material as ordinarily furnished. Of 



