Temperature and the Serenity of the Sky. 155 



1835-45, 48°"1 *. Now on referring to the curves (figs. 2 and 3), 

 it will be seen that, notwithstanding this great difference, the 

 eflFects which had been previously remarked in the Dublin and 

 Greenwich curves are still in each instance very apparent. That 

 is to say, the mean temperature is above the average in both 

 curves, (1) at the period of new moon, (2) at first quarter, and 

 (3) before last quarter. It is below the average, (1) between new 

 moon and first quarter, (2) before and after full moon, and (3) at 

 and after last quarter. 



These six periods of high and low temperature should there- 

 fore be conspicuous in the curve of twenty years' mean tempe- 

 rature. And this is also found to be the case ; more particularly 

 at new moon, and for five or six days before and after it, 



3. It was early perceived, however, and would seem to be a 

 point of much importance, that a peculiar influence, which ap- 

 pears to be exerted about the middle of the lunation, advances 

 or retards effects which nevertheless still show themselves plainly 

 a day or so earlier or later in the curves. The rise in tempera- 

 ture at full moon (see a in Plate), which so often takes place for 

 a short time in what is otherwise a period of low mean tempera- 

 ture, is an instance of this. On referring to the curves it will 

 be seen that, in the instance when the rise occurred before full 

 moon, the mean temperature is highest before first quarter; 

 when it was after full moon, the rise occurs after first quarter. 

 This movement may possibly prove to be a law rather than a 

 disturbance ; but whether this turn out so or no, it is evident 

 that the action would in a series of years obliterate very regular 

 and systematic lines in the curves. For this reason, and also 

 because different effects attend upon lunar influence at different 

 seasons of the year, it appeared to be the better plan to form 

 curves of mean temperature for individual months. The remark- 

 able manner in which the low temperature in November last, 

 whilst defying the expectations of meteorologists, followed the 

 outline of the curves of annual mean temperatui-e at Greenwich, 

 hastened the commencement of the scheme. 



4. The long-continued frost of that month, it will be remem- 

 bered, set in about the 9th day of November, or the fourth day 

 after new moon. It was interrupted by a thaw of three days' 

 continuance at first quarter, which raised the mean temperature 

 to 42°; after which the cold set in with renewed intensity, till 

 on the 19th day of the month — two days before full moon — 

 it was 13 degrees below the average of that day (of the month) 



* The average mean temperature of the curves is somewhat higher in 

 both instances. This is due to commencing the meteorological year with 

 November, and to the omission of the means of the observations in the 

 octant columns. 



M2 



