422 Mr. F. H. Pitcher on Effects of Temperature and 



obtained. The incomplete series of tests obtained in this 

 manner were, however, of interest as a verification of the 

 method subsequently adopted. 



In the meantime, while awaiting the production of suitable 

 tubes, it was thought that interesting results might be ob- 

 tained by heating the iron wire with an electric current 

 passed through the wire itself, and deducing its mean tem- 

 perature from its resistance, with the aid of the formula 

 verified by Prof. Callendar (Phil. Trans., A. 1887, p. 225) by 

 the direct comparison of platinum and iron wires. The 

 objection to this method of heating is that the wire is cir- 

 cularly magnetized by the heating current, and that it is 

 necessary to disentangle the effects of the temperature change 

 and of the circular magnetization on the longitudinal mag- 

 netization of the specimen. The effect of the circular field 

 itself, however, is not without interest* 



In order to disentangle these effects, three separate series of 

 observations were taken for the same range of current or circu- 

 lar field, (1) in a very high vacuum ; (2) in air at atmospheric 

 pressure; (3) in a current of water. In case (1) a current 

 of 16 amperes sufficed to heat the wire above its critical 

 temperature ; in case (2) the highest temperature was 

 400° C. : in case (3) the heating effect was practically 

 negligible. 



Methods of Measurement Adopted. 



The iron wire specimen was magnetized by means of a 

 specially constructed solenoid, and the intensity of mag- 

 netization I at any time was observed by means of the 

 deflexion of a magnetometer, the direct effect of the solenoid 

 being very carefully compensated by means of a balancing 

 coil in the usual manner. The broadside- on position was 

 adopted for the test in preference to the vertical or the end- 

 on position, as it had been found by preliminary tests that, 

 if the distance of the specimen from the magnetometer were 

 suitably adjusted, the broadside-on method agreed much 

 more closely with ballistic tests of the same specimen than 

 either of the more usual positions. The value of the Earth's 

 field H was repeatedly determined by the aid of a Kohlrausch 

 variometer. 



The deflexion of the magnetometer was observed by means 

 of a telescope and a metre-scale of milk-glass very accurately 

 divided. The magnetometer was also provided with suitable 

 galvanometer-coils, so that readings of current and resistance 

 could be taken on the same scale. The scale of the gal- 

 vanometer was carefullv calibrated throughout, and all the 



