PHIL0S01»HICAL WRITINGS OF THE LATE DR. DALTOX. 25 



C and D ; lastly throu|^h C draw PG, DQ, and the thing is 

 done." 



Demonstration. For PM : MN :: P6 : 60 :: PB : BC; 

 and since the triangles BCD, GCQ, are similar DC : CQ :: 

 BC : CO :: bo : oT :: MN : NO ; the given ratio. 



Limitation, ^' If either of the parallels BD, oC, fall 

 without the circle, the problem is impossible." 



In the Ladies' Diary and Supplement for 1794, Mr. 

 Dalton answers three of the queries and four of the mathema- 

 tical questions, including the prize. The Jirst query in the 

 Diary was proposed by himself, and requires " the cause of 

 the mist which is sometimes observable in a calm evening to 

 hover over meadows and rivers." His explanation of the 

 phenomenon, which agrees with that furnished by many other 

 correspondents, is to the effect that the mist "is only 

 observed when the air has suffered a sudden change of tem- 

 perature from heat to cold. It is found from experience that 

 warm air will hold more water in solution than cold air; 

 therefore when the air is suddenly cooled, which sometimes 

 happens in an evening, the water, being then much warmer 

 than the air, evaporates pretty copiously at the surface, but 

 is no sooner carried up a little into the cold air, than it is 

 precipitated again in the form of mist, and occasions the 

 phenomenon." The Supplement contains his theory of perio- 

 dical winds, which is followed by an editorial notice of the 

 publication of his essays on meteorology, and his own opinion 

 on the effects of burning charcoal on steel. On the former 

 subject he observes, that " the unequal capacities of land and 

 water for receiving heat are undoubtedly the causes of these 

 phenomena. Land receives more heat than water, or rather 

 perhaps is heated to a greater degree. Independently, there- 

 fore, of the general or trade winds, the winds should blow 

 towards the centre of the [West India] Islands in the day, 

 and from the centre in the night ; consequently, on the 



