﻿448 Prof. Threlfall on the Scattering 



effect is due to scattering. Tins, however, appears to be most 

 probably the case, as the following experiments will show : — 



Three typical samples prepared by Prof. Liversidge were 

 examined. The first sample had been made up about two 

 months before I examined it. The quantity of gold was so 

 large that the precipitate had partly subsided ; and the 

 solution was so dark that it was practically opaque in thick- 

 nesses of more than a few centimetres. The absorption 

 spectrum was observed with the help of a wedge-shaped 

 bottle. The absorption was chiefly in the green and greenish 

 blue. At about the limit of transparency practically all the 

 spectrum disappeared except the orange-red. The opacity 

 was such that no satisfactory observations could be made on 

 the scattered light. 



The second sample had a magnificent pink colour, very like 

 a dilute solution of fluorescein, but of course without the 

 fluorescence of that substance. The colour was exceedingly 

 brilliant, and having only had experience of the colour pro- 

 duced by precipitates of gums in water, for a long time 

 I imagined that some coloured substance had been formed. 

 This specimen was described by Prof. Liversidge (loc. cit. 

 p. 406). It was prepared in November 1889, and from the 

 description of its colour cannot have changed appreciably 

 since that date. The reducing substance was phosphorus in 

 chloroform, and a smell of chloroform was still marked. The 

 solution of gold chloride was of about the strength mentioned. 

 The absorption spectrum was observed in a tube 69*3 centim. 

 long. The absorption was not nearly so sharp as in the case 

 of, say, chlorophyll, but was wonderfully sharp for a " sky n 

 colour. The absorption extended from the yellow-green to 

 the blue-green ; there also appeared to be a slight absorption 

 in the blue ; but it was very slight. The scattered light was 

 easily observed in this case, especially when a Nicol prism 

 was used in conjunction with the spectroscope. It was thus 

 discovered that the scattered light was practically green, and 

 of about the proper apparent intensity. A line was drawn on 

 a board so as to indicate a directicn at right angles to the 

 direction of the incident light ; and another line was drawn 

 at 120 degrees to the same direction. By means of these 

 lines the position for maximum polarization could be suffi- 

 ciently closely observed. There can be no question as to the 

 angle being much closer to 90° — even if it were not exactly 

 90°, as the observation seemed to show — than to 120°. I 

 cannot say within what limits of angle the position was 

 perpendicular to the direction of the incident light. 



The third sample I examined had a full purple colour, and 



