1868.] 437 [Chase. 



latitude, and within the influence of the Gulf Stream, therefore 

 the lunar modifications of the climate, especially when those 

 modifications are estimated merely from their equatorial means, 

 are comparatively unimportant ; the four lunar quarters, each 

 embracing an entire interval between a spring and a neap tide, 

 are precisely the divisions which, when the periods are suffi- 

 ciently extended, would eliminate most of the remaining evi- 

 dences of lunar disturbance. Still, if we examine the half- 

 cycles, or ten-year periods, we may observe that the half- 

 months of lunar conjunction present a rainfall alternately 

 greater and less than the half-months of opposition, a fact 

 which suggests a possible dependence on the alternately pre- 

 dominating north and south latitude. 



It is impossible, from the tables, to determine the extent of 

 such a dependence. Mr. Dines, who has the necessary mate- 

 rials at his command, ma}^ perhaps deem the subject worthy of 

 further investigation. The division of the month'may be easily 

 altered so as to correspond more nearly with the times of pre- 

 sumable maximum and minimum disturbance. If we regard 

 the day of each change of phase as the middle day of a week, 

 (counting the half-sum of the 5th and 12th days in the first 

 quarter, and the half-sura of the 20th and 27th days in the third 

 quarter), the seveii-da3\s' aggregates in Table I, and in the two 

 summaries on pp. 136-t (loc. cif.), will furnish the following 

 results : 



been led to the investigation by Howard's remarks " on a cycle of eigh- 

 teen years in the mean annual height of the barometer in the climate of 

 London." Clim. of London, 2d ed., v. 1, p. 172; Phil. Trans., 1841, p. 

 277, seq. See also Zenger's discussion of the mean annual temperature, 

 as affected by the revolution of the moon's nodes and apsides. Phil. 

 Mag., V. 35, June, 1868. 



