( 214 ) 



C 



Q — 

 La- 



Fig. 1. 



unequally distributed, particulars concerning the dis- 

 tribution would be revealed by the spectroheliograph 

 through the refracting power of the vapour, rather 

 than through its absorbing and emitting power. This 

 expectation could be put to the test. 



As an equipment for the study of anomalous dis- 

 persion phenomena in sodium vapour, exactly similar 

 to the one described in my paper on "Arbitrary dis- 

 tribution of light in dispersion bands" 1 ), had already 

 been procured for the solar observatory by Professor 

 Hale, the experiments were readily made, thanks to 

 the laboratory facilities available on the mountain. 



The apparatus consists of a wide nickel tube, 60 cm. 

 long, the middle part of which is placed in an electric 

 furnace, while the projecting ends are cooled by jackets 

 with flowing water. The tube contains a few grammes 

 of sodium and is permanently connected to a Geryk- 

 pump to remove the air and the gases which escape 

 from the sodium during the first stages of the heating 

 process. An arrangement is provided for, by which 

 density gradients of various known directions and 

 arbitrary magnitude may be produced in the sodium 

 vapour. 



Sunlight coming from the 60 feet mirror M (fig. 1) 

 of the Snow telescope 2 ) passes through the tube T on 

 its way to the slit .$ of the spectroheliograph. The 

 distance between 7' and S is about 560 cm. A lens 

 L x gives an image of the sun near the middle of the 

 tube T. P is a diaphragm with an adjustable slit, of 

 which the lens L, projects an image in the plane of 

 the diaphragm Q. Just behind the latter is a lens L t ; 

 in combination with L t this forms an image of a 

 section of the tube in the plane of the slit S of the 

 spectroheliograph. In this image (fig. 2) the rectangular 

 windows of the caps of the tube will of course come 

 out with somewhat blurred edges, as only a section 

 lying somewhere between the caps would show sharp. 

 In A and B are projected the narrow nickel tubes '), 



i) Proc. Roy Acad. Amsterdam, IX, p. 343. (1906). 



2 ) Described in: Contributions from the Solar Observatory 

 Mt. Wilson, Gal., Nos. 2 and 4. 



3) See description in: Proc. Roy. Acad. Amsterdam, IX, p. 345. 



