122 



BRIDGMAN. 



obtained from micrometer measurements of its dimensions, was 59 X 

 10""^ ohms per cm. cube. 



Neodymium. This, as well as the lanthanum, I owe to the kindness 

 of Professor Baskerville. Professor Saunders was kind enough to 

 make a spectroscopic analysis of this also. ■ He found a large amount 

 of Mg, a little Si and La, a trace of Ca, no Ba, and nothing else recog- 

 nizable. He tested for the rare earths Ce, Pr, Y, Yt, Lu, Dy, Er. 

 The form and method of preparation of the specimen was essentially 

 the same as that of lanthanum. It was extruded into wire 0.020 

 inches in diameter at 450°. The extrusion was materially easier than 

 that of lanthanum. The w're is not so stiff, and may be straightened 

 after extrusion without fear of breaking. Nevertheless it obviously 

 belongs to the metals with high elastic constants and high elast'c limit. 

 The specimen used for the measurements was 7.1 cm. in length; the 

 manner of attachment of the connections was the same as \\'ith 

 lanthanum. The greater length of the specimen, and perhaps greater 

 skill in handling it, led to much more regular results. 



A preliminary measurement of the temperature coefficient was 

 made at 0°, 50°, and 95°. Within this range the relation between 

 temperature and resistance was found to be linear, and the coefficient 

 was 0.000799. This is extraordinarily low, much lower than for 

 lanthanum even, and it did not seem worth while to expend a great 

 deal of time on the pressure measurements. 



Two runs were made for the pressure coefficient, at 0° and 50°. 

 The results were rather regular. There was no difference between 

 readings with increasing and decreasing pressure, and the zero was well 

 recovered. The maximum departure of an}- single point from a 

 smooth curve was 2.6% of the total effect at both 0° and 50°. The 

 numerical results are shown in Table XYI. The ^-alues are quite 



TABLE XVI. 



Neodymium. 



