﻿454 Prof. Threlfall on the Scattering 



decomposed. (Professor Wright, of Yale, warned me in 

 1889 that iron mirrors are not easily made.) The experi- 

 ment was perforce interrupted for three days, and on turning 

 on the coil on the fourth day a smart explosion resulted, 

 bringing the observations to a conclusion by scattering every 

 part of the apparatus, the writer having a rather lucky escape. 

 The electrodes were picked up and found to be perfectly clean 

 and bright. After consideration I decided not to pursue the 

 matter further, first because I am not quite sure of the 

 importance of an investigation of iron in the present state of 

 electromagnetic theory, and secondly because the investigation 

 can be easily made by anyone who has a suitable hydrocarbon, 

 whereas I can only hope to obtain inconclusive results with 

 the oils at my disposal. 



I will add a note on producing oxygen-free atmospheres. 

 If hydrogen is employed the difficulty is to get rid of oxygen 

 from the water and acid used to act upon the zinc. Several 

 ways of absorbing small quantities of oxygen are known and 

 dealt with in a paper by me, " On the Preparation of Pure 

 Nitrogen and Attempts to Condense it," Phil. Mag. January 

 1893. 



The most convenient way of obtaining large quantities of 

 hydrogen free from oxygen is to half fill a very large flask 

 with granulated zinc and keep this covered with a solution of 

 chromous chloride — the l< liquide brute " of Recoura. Hydro- 

 chloric acid can be added fearlessly by means of an ordinary 

 safety funnel ; for any air carried down is instantly deprived 

 of its oxygen by the blue liquid. If the flask be permanently 

 sealed to a potash wash-bottle and drying-tubes, a means of 

 procuring dry hydrogen with traces of nitrogen is instantly 

 and continually available. The blue liquid does not last for 

 ever, however, though if a little acid be added from time to 

 time its life is sufficiently prolonged. Following Recoura in 

 the paper quoted, I recommended that the potassium dichro- 

 mate should be finely ground in the chromous chloride 

 preparation. I now find that if it is dissolved in the hydro- 

 chloric acid before the latter is added to the zinc, it does just 

 as well and saves a good deal of trouble. 



To sum up : — 



1. The scattered light from gold and platinum particles 

 behaves like light scattered from particles of gum mastic or 

 milk so far as the polarization phenomena are concerned. 



2. The same remark applies to iron, or iron oxide, or 

 carbide, whichever it was that I examined. I rather think, 

 on the whole, that the sky was iron simply. 





