with a Liquid Mercury Anticathode \ 421 



between A and B, and B and the third part G which con- 

 tained the anticathode. C was made of a brass tube of about 

 10 mm. diameter and closed by a copper rod which fitted in 

 the tube. The copper rod was sealed in C with sealing-wax. 

 The upper end of the rod was amalgamated with mercury, 

 and a small quantity of mercury had been poured in the 

 tube C, so as to cover the flat end of the rod with a thin film 

 of mercury which formed the anticathode. A cubical metal 

 block surrounded the tube C at the height of the anticathode 

 surface. The X-rays left the bulb through an opening which 

 was drilled through the block and the tube. This opening- 

 was covered by a thin mica sheet. A very efficient cooling 

 was produced by two water-jackets ivj. In some of the 

 experiments the whole, bottom part C was made of pyrex 

 glass, and filled partly with mercury. The cooling was in 

 this case not very efficient on account of the smaller thermal 

 conductivity of the glass. The final results have all been 

 obtained with the first-mentioned arrangement. A cylin- 

 drical brass vessel was waxed on the top of the glass bottle 

 which served as discharge chamber. The aluminium cathode 

 was fixed on this vessel, so that its surface was near the 

 lower end of the neck of the bottle. The vessel was filled 

 with water in order to cool the joint with the glass. 



A second bulb of the same kind but with a copper anti- 

 cathode was used in order to get the copper lines on the 

 spectrograms. These lines have been used as standards for 

 measuring the Ho- lines. 



3. Description of the Spectrometer (fig. 2). 



Only the data will be given here which are necessary for 

 the discussion of the results. The mechanical details of the 

 apparatus will be described in a separate paper. 



The method of the revolving crystal was used in the 

 present work. The arrangement of the essential parts of 

 the spectrometer is shown in the diagram. The width of the 

 slit S x was measurable to '0L mm. A width of *05 mm. was 

 generally used. Photographs obtained with '02 mm. did not 

 show more details. The second slit S 2 served to limit, the 

 primary beam of X-rays. The crystal-holder was made so 

 as to allow the crystal surface to be adjusted parallel to the 

 slit and to be brought within a suitably small distance from 

 the axis of revolution. The measurement of this distance 

 will be described later on. The distance between the <\\\ 

 8] and the axis of revolution was about 5 cm. The length 

 of the perpendicular line from the axis to the plate was 



