THE MAGNETIC RADIOMETER l8l 



not required. A thin strip of zinc I mm. in breadth and 

 10 cm. in length was therefore used as the sensitive ele- 

 ment, zinc wire not being available. It was cut and filed 

 sufficiently thin ; the two ends of this strip were soldered 

 to thicker pieces of brass in which large holes were bored, 

 filled up with ivory. A small hole was bored in the ivory 

 for the insertion of the upper hook which attached the strip 

 to the magnetic lever ; the second piece of brass soldered to 

 the lower end was clamped between two jaws faced with 

 non-conducting sheets of mica ; a pin was passed through 

 the narrow hole in the ivory which plugged the lower piece 

 of brass to make the clamping more secure. The object of 

 the thermal insulation of the sensitive strip is to prevent 

 its receiving any heat by conduction from the rest of the 

 apparatus. 



The strip was surrounded by a wooden tube with a 

 narrow slit for the passage of the spectral rays. The slit was 

 open through a length of 10 cm., the brass piece to which the 

 strip was soldered being protected from radiation. A very 

 thin piece of mica covered the slit, to prevent air-currents 

 getting access to the narrow chamber in which the sensitive 

 strip was enclosed. This exceedingly thin piece of mica was 

 found to have no effect in obstructing radiation. The whole 

 apparatus was placed inside a larger wooden box covered 

 with non-conducting felt. A round hole in front of the box 

 (covered with a piece of thin glass) allowed the passage 

 of the spot of light reflected from the mirror attached to the 

 suspended magnetic needle. A second thin strip of mica 

 closed the slit behind the box through which the radiation 

 reached the sensitive strip. 



It is necessary to give a detailed account of precautions 

 to be observed, because the instrument is so sensitive that 

 it detects the slightest difference in the temperature of 

 different portions of the same mass of air. Gases are highly 

 non-conducting, and their temperature hardly attains a 

 perfect uniformity. In these circumstances it is best to 

 take precautions against contact of the strip with the 



