348 On the influence of the Moon [July, 



of the degree of moisture. For this reason I took only the first and last 

 quarters of the year, when the weather is usually dry, and found the means 

 of the different years (see table No. 4). From this it will be seen that the 

 days of highest dew-point in the winter half of the year are the 4th and 

 5th, before the new moon, and the 10th and 11th, after. This was so 

 near a coincidence with two of the maximum days of rain, (viz. the 3rd 

 before, and the 12th, after,) that little doubt could be entertained of the 

 one being caused by the other. That neither the other two days, (viz. 

 the 9th, before, and the 5th, after,) were maxima might be accounted for 

 by peculiar circumstances. At this stage of the inquiry I was led to 

 attempt to account for the phenomena by the following considerations : 



1st. By the united testimony of every observer, the quantity of mois- 

 ture in the air and the rain-fall become less, as we recede from the great 

 eastern ocean. Thus if we could obtain the mean dew-pointfor every degree 

 of longitude between Delhi and Dacca, the result would exhibit almost 

 as regular an increase as in a list of temperatures between London and 

 Algiers. I was aware too of the great increase of dew-point here when- 

 ever the wind came from the east, and that a continuance of it was 

 usually followed by rain. I could not, therefore, but believe that the 

 force of attraction of the moon as well as of the sun excited an influence 

 over the aerial currents either in modifying their direction or changing 

 it entirely. Mr. Daniell remarks the excess of dew-point when the 

 wind blows from the Atlantic (he is speaking of the climate of England), 

 and the force of attraction of the moon is stated by D'Alembert to be 

 such as would create a westerly current of eight feet in a second, (see 

 Robison's Mechan. Phil.) But to render this force more apparent, we 

 must have recourse to another consideration. 



2ndly. The principal cause by which the air is affected is by the heat- 

 ing power of the sun, which expands a column of it ^th part for 

 each degree of Fahrenheit. Upon the ocean the heat is counteracted by 

 constant evaporation ; consequently,when a column of air, resting upon a 

 surface of dry land, is heated by the sun, it becomes expanded, and of 

 less specific gravity than an adjoining column in contact with the sea. 

 Hence, as the heavier fluid will press upon and displace the lighter, a 

 current flows in from the sea to the land. This diurnal phenomenon may 

 be observed on almost every tropical coast. We have an annual instance 

 of it in the great heats previous to the summer solstice, and the winds 

 that follow them from every quarter of the ocean, the S. W. the S. E. 

 and E., bringing with them abundant rain. This current must be 

 strongest at the time of maximum heat of the day, and if we suppose 



