CHAP. 2. 16. COLOURS OF CLOUDS. 87 



former is however probably the case ; for dif- 

 ferent clouds in nearly the same angular posi- 

 tion with respect to the sun show different 

 colours at the same time. But the colours re- 

 fracted by the haze are very various. Sometimes 

 the tints in the twilight haze come on so suddenly 

 and are so circumscribed as to induce a belief that 

 very sudden and partial changes take place in 

 the atmosphere at eventide ; which may perhaps 

 be somehow connected with the formation of 

 Dew. It is doubtless the falling Dew which 

 refracts the colours in general, which are varied 

 by the position of the parts in which they are 



or fit of dilation and brightness of the Star seemed whitish, and 

 the next or alternating twinkling seemed red. I have elsewhere 

 described this appearance more minutely, and have attributed 

 it to some wavy motion in the upper part of our atmosphere, 

 whereby, being imperfectly prismatic, it produced different 

 colours according to the relative inclination of the undulating 

 surface. At other times I have thought it might arise from 

 some change in the Star itself; but the former proposition is 

 much more probable ; and what corroborates the idea that it 

 is atmospheric, is that it is not always equally conspicuous on 

 all kinds of nights. I should like to know how the East wind 

 affects this phenomenon, as when the wind is from that quarter, 

 celestial objects seem to dance about so in the field of the 

 telescope, that no good observations can be made. Refer to 

 Phil. Magazine for 1817? vol. xlix. p. 453, and to the Monthly 

 Mag. for Jan. 1823, vol. liv. 486. 



