of Light by Potassium Vapour. 293 



The photographs were taken under the following con- 

 ditions: the potassium having been placed in the tube and 

 the tube exhausted, a current was passed through the heating 

 coil. Initially, the central fringe was horizontal and the others 

 slightly curved in opposite directions above and below this 

 central fringe, the separation between the fringes being 

 greatest at the red end. When the temperature increased to 

 about 300° 0. the fringes near the red absorption line 7675 

 showed a change in direction. The absorption near the blue 

 line 4045 was too slight at this stage for any change in 

 direction of the fringes to be noticeable at the bine end of 

 the visible spectrum. 



The photographs were taken when the reading of the 

 voltmeter remained steady. As the temperature increased 

 and the radiation from the asbestos covering of the heating- 

 coil became greater, it was found increasingly difficult to 

 obtain steady temperatures. At first it seemed only possible 

 to obtain consecutive photographs at low temperatures, the 

 amount of dispersion near the blue line being very small. 

 Several series of three or four sharply defined photographs 

 were obtained, but in these the slope of the fringes is very 

 slight except near the absorption lines, and measurements 

 of points where the fringes cross a horizontal line are 

 uncertain ; also the temperature range is too small to test 

 the constancy of the ratio a\ja % . By widening the slit and 

 reducing the exposures to about 10 seconds, this difficulty 

 was eventually overcome. 



The plates used were W ratten k Wainwright's Pan- 

 chromatic A, but it was not found possible to detect the 

 fringes for higher wave-lengths than about 7670. The 

 fringes on the other side of the red line were therefore never 

 obtained on the photographs. 



The absorption region to the right of the red line began 

 to appear at an early stage. The fine dark lines of which it 

 is made up were easily visible through the spectroscope and 

 appear on some of the negatives. In the photographs corre- 

 sponding to lower temperatures the path of the fringes 

 through this region can be traced, but for higher temperatures 

 this is impossible. The edge of this region on the side 

 furthest from the red absorption line is sharply defined, and 

 for all temperatures the fringes are quite clear down to 

 wave-length 6000. But the range of measurement at this end 

 of the spectrum was curtailed by the presence of the sodium 

 lines. It was found impossible to eliminate all traces ot 

 sodium, and as at high temperatures the dispersion at these 

 lines was appreciable measurements were not continued 

 beyond this point. 



