326 



Dr. T. Ewan on the Absorption-Spectra of 



of layer was chosen so that the absorptions to be measured 

 should as often as possible fall between these limits. Three 

 cells were used — viz. a Schulz cell, the glass cube of which 

 had a thickness of 1*006 centim., and two troughs,. 7*64 and 

 22' 86 centim. long respectively. When using these troughs 

 the liquid only filled the lower half, the meniscus being placed 

 at the level of the division between the two halves of the 

 double slit. As, however, this produces a black band, about 

 2 millim. broad, between the two spectra, thereby greatly 

 increasing the difficulty of comparing their brightness, a flint- 

 glass rhomb was placed between the trough and the slit, in 

 the same way as in Hiifner's spectro-photometer*. The 

 arrangement is shown in fig. 2. 



Fig-. 2. 



The ends of the trough were provided with screens of 

 blackened card of sufficient width to prevent reflexions from 

 the sides of the trough or from the meniscus. The source of 

 light was, as before, the Argand burner. The measurements 

 were made in a dark room, and extraneous light shut off from 

 the apparatus as much as possible by screens, and from the 

 eye of the observer by a piece of black velvet which covered 

 the eyepiece of the observing-telescope. The light from the 

 Argand burner was concentrated by means of two lenses on a 

 hole, about T 2 g inch diameter, in a black screen, and after 

 passing this was made parallel by a lens of about 15 inches 

 focal distance. The light thus obtained was rather faint, but 

 it was not found possible to obtain reliable results otherwise. 

 The salts examined were carefully purified and the solutions 

 made with twice distilled water. They were filtered imme- 

 diately before being examined through fine Swedish filter- 

 paper. All the solutions were analysed after being filtered, 

 as the filtration made a small difference in the concentration. 

 The analyses were made by the very convenient method — 

 titration, with thiosulpbate solution, after adding excess of 

 potassium iodide — described by Westmoreland f. The con- 



* Zeit. Phys. Chem. iii. p. 562 (1889). 

 t Journ. Soc. Chem. Ind. 1886, p. 48. 



