118 FRAGMENTS OF SCIENCE. 



Roscoe has noticed several striking cases of a similar 

 kind. In a very remarkable paper the late Principal 

 Forbes showed that steam issuing from the safety-valve 

 of a locomotive, when favourably observed, exhibits at 

 a certain stage of its condensation the colours of the 

 sky. It is blue by reflected light, and orange or red by 

 transmitted light. The same effect, as pointed out by 

 Groethe, is to some extent exhibited by peat-smoke. 

 More than ten years ago, I amused myself by observ- 

 ing, on a calm day at Killarney, the straight smoke- 

 columns rising from the cabin-chimneys. It was easy 

 to project the lower portion of a column against a dark 

 pine, and its upper portion against a bright cloud. The 

 smoke in the former case was blue, being seen mainly 

 by reflected light; in the latter case it was reddish, 

 being seen mainly by transmitted light. Such smoke 

 was not in exactly the condition to give us the glow 

 of the Alps, but it was a step in this direction. 

 Briicke's fine precipitate above referred to looks 

 yellowish by transmitted light ; but, by duly strengthen- 

 ing the precipitate, you may render the white light of 

 noon as ruby-coloured as the sun, when seen through 

 Liverpool smoke, or upon Alpine horizons. I do not, 

 however, point to the gross smoke arising from coal as 

 an illustration of the action of small particles, because 

 such smoke soon absorbs and destroys the waves of blue, 

 instead of sending them to the eyes of the observer. 



These multifarious facts, and numberless others 

 which cannot now be referred to, are explained by 

 reference to the single principle, that, where the scatter- 

 ing particles are small in comparison to the ethereal 

 waves, we have in the reflected light a greater propor- 

 tion of the smaller waves, and in the transmitted light 

 a greater proportion of the larger waves, than existed 

 in the original white light. The consequence, as 



