126 BRIDGMAN. 



and there were numerous fairly large pores. (There were visible pores 

 in the first sample also). It need not be anticipated that the pores 

 cause any error in the pressure coefficient, for the transmitting liquid 

 freely penetrates the pores and transmits pressure uniformly to all 

 parts; there is never any permanent change of dimensions after 

 an application of pressure. The second sample was of approximately 

 the same dimensions and was treated in the same way as the first. 

 Two sets of readings were made with this second sample, at 0° and 95°. 

 The pressure coefficient is negative in sign, as it was for the first 

 sample, but the numerical values are somewhat different. At 0° 

 the resistance decreases by 10.1% under 12000 kg., and at 95° by 

 15.3%. The change is not linear with pressure, but the coefficient 

 becomes larger at the higher pressures, which is the opposite of the 

 normal behavior of the first sample. The temperature coefficient 

 of this sample between 0° and 95° was + 0.0000615, about half as 

 large as that of the first sample. 



In spite of the very marked differences these two samples agree 

 much more nearly in their pressure coefficients than they do in their 

 temperature coefficients. This agrees with previous experience, that 

 in general the temperature coefficient is much more susceptible to im- 

 purity than the pressure coefficient. We may expect, therefore, that 

 the pressure coefficient of resistance of pure silicon will be found to be 

 negative, and of the order of —0.000012, pressure being expressed in 

 kg/cm-. Compared with most metals, this coefficient is high, being 

 about the same as that of lead. 



Black Phosphorus. Runs were made on two samples of this 

 substance. The first was from the same piece as that which gave 

 the values for the specific resistance and temperature coefficient of 

 resistance already published. ^° The method of formation and some 

 of the other properties have also been described. During the six 

 years since the previous measurements, this specimen has been kept 

 in a glass bottle, closed with a cork stopper and sealed with paraffine. 

 The protection from the action of the air was not perfect, however, 

 because the phosphorus had become covered with a layer of moisture. 

 This moisture is probably due to slow oxidation of the phosphorus 

 in the air. The result of oxidation is the formation of phosphoric 

 acid, which is well known to be very hygroscopic, and therefore rapidly 

 absorbs moisture from the air. An attempt was made to remove the 

 acid from the sample by boiling it with water for a number of hours, 

 and then heating in vacuum for a number of hours in addition. 



The specimen previously used was a cylinder about 0.5 inches in 



