PRESSURE ON RESISTANCE OF METALS. 615 



For the initial pressure coefficient at 0° the above deviation curves 

 give by graphical construction — O.OslQS, against — O.O5I77 found by 

 Lisell.'^ He does not give the temperature coefficient of his platinum, 

 but chemical analysis showed an impurity of 0.28% iridium. We will 

 see later that the discrepancy is very probably due to impurity. 

 Laf ay ^ gives for "pure" platinum the pressure coefficient O.O5I86; 

 he found the relation between pressure and resistance linear up to 

 4000 kg. 



Before this run with Heraeus platinum, a complete series of runs 

 was made with a specimen from Baker and Co., their purest. This 

 was subjected to preliminary treatment in all details like that of pal- 

 ladium, and the results were of the same regularity as those for palla- 

 dium. The pressure coefficient was found to decrease linearly with 

 temperature from O.O5I766 at 0° to O.O5I742 at 100°. The deviation 

 curves were not symmetrical, and have the same distinct progression 

 with temperature of the pressure of maximum deviation as have those 

 of nickel. But a low value for the mean temperature coefficient was 

 found, 0.003466, showing distinctly large impurities. It is to be 

 noticed that the initial pressure coefficient at 0° is almost exactly the 

 same as that found by Lisell. This makes it almost certain that the 

 lowness of his pressure coefficient is due to the 0.28% of iridium. The 

 fact that the deviation curves for impure platinum show a marked 

 progression of the maximum, whereas those of pure platinum do not, 

 suggests that the progression of the maximum found for nickel may be 

 a spurious effect due to impurity. 



The general character of the results is exactly the leverse of what 

 we should be tempted to call normal. If the resistance-pressure curves 

 are scaled to the same initial resistance for all temperatures, the slope 

 and curvature are both less at higher temperatures. 



Molybdenum. This was obtained from the General Electric Co. 

 through the kindness of Dr. W. D. Coolidge. It was in the form of 

 bare wire 0.0011 inch diameter. It was seasoned by a number of 

 excursions between 0° and 130°, and was then wound bare on a bone 

 core. The resistance at 0° was 83.8 ohms. Some little trouble was 

 found in making a suitable connection; silver solder will not stick 

 to it. It may be readily attached to platinum by arcing in hydrogen, 

 which is the method employed by the General Electric Co. But this 

 does not make a perfectly sharp contact, there being a small region 

 of imperfect contact where the fine molybdenum wire leaves tangen- 

 tially the surface of the larger platinum wire. It is very difficult to 

 make a butt joint without the use of special devices. It seemed to me 



