14 PHYSICAL SCIENCE BK. i 



same nature as the halo may often be witnessed 

 in baths, because the atmosphere is thick and dark : 

 it is most frequent when the wind is in the south, 

 when the air is heaviest and most dense. 



5 Halos sometimes are dissolved gradually and 

 fade away, sometimes they are broken up on 

 one side. In the latter case seafaring men look 

 for wind in the direction in which the circle of 

 the crown has been broken. If the parting is on 

 the north, there will be a north wind, if on the 

 west, zephyrs will follow. This is a proof that 

 these crowns are formed in the region of the sky 

 in which the winds are usually formed. The 

 upper regions of air have no crowns because they 



have no winds either. An additional proof of the 

 connection of winds and halos is afforded by the 

 fact that the halo is never formed unless the atmo- 

 sphere is at rest, and the wind, as it were, inactive. 

 Under other circumstances it is not usually observed. 

 The atmosphere when it is at rest may be 

 fashioned to any pattern by being driven or drawn in 

 any direction. But when it is in motion, light cannot 

 even strike it. It takes no shape and offers no 

 resistance, because the part first affected is always 



7 dissipated by the motion. Therefore it is that no 

 heavenly body can ever be surrounded by a figure of 

 the kind referred to unless when the atmosphere is 

 dense and motionless, and so preserves the ray of 

 round light that strikes upon it. Nor is it without 

 good reason. Recollect the analogy mentioned a 

 little ago. A pebble thrown into a pond or lake or 

 any other circumscribed piece of water produces 

 innumerable circles ; but it has not the same effect 

 if thrown into a river. And why so ? Because in 

 the latter case the water as it hurries on prevents 



