742 J. A. BROUN ON THE LUNAR DIURNAL VARIATION OF 



19. 4:th, While the ]unar diurnal variation changes the hours of maxima and 

 minima more gradually than the solar diurnal variation, it also makes the 

 greatest changes at different times. Thus the solar diurnal variation changes 

 completely during the month of March, or from February to April, while the 

 lunar diurnal variation makes the greatest change from April to May. The 

 second change which happens for the sun between September and November 

 occurs earlier, or between September and October for the moon. 



Range of the Mean Lunar Diurnal Variation. 



20. When we examine the range of the variation in the different months of 

 the year, we find : — 



5th, That the range is greatest in January, and least in May and October ; 

 the arc, including the mean diurnal variation for January, from eleven years' 

 observation, being nearly 0'5 (= 30"), while in May the range was 0'18 

 (= 10"-8), in October 0'14 (= 8"-4), and in July 0'-26 (= 15"'6). 



21. It has been shown in a paper already cited/'' that the lunar diurnal varia- 

 tion is sometimes as large as the solar diurnal variation, amounting in December 

 and January sometimes to upwards of five minutes of arc (5'), which, allowing 

 for the greater inclination of the needle (or the smaller horizontal force), would 

 be equivalent to about 12' in England. But this great oscillation, which some- 

 times occurs within 24 hours, is subjected to different laws of variation which, 

 when the mean for a week only is taken, diminish the range so much that, for 

 example, in the lunation 16th December 1858 to 12th January 1859,t the 

 greatest range for a week's observations is reduced to 2'7, while the range for 

 the whole lunations is less than half that amount. When the mean diurnal 

 variation is derived from all the lunations occurring principally in January 

 during eleven years, the range is still further reduced to 0'*5 nearly. This is 

 partly due to the fact that the lunar action does not appear to be equally 

 powerful in the same month in different years.! 



22. Qth, The ranges of the mean lunar and solar diurnal variations thus obey 

 different laws relatively to the epoch of the year ; the range for the former in 

 January being nearly double that in any month from May to September, while 

 the range of the latter (the solar diurnal variation) in January is little more 

 than half that for August. 



23. Although it would be difficult to prove from the ranges of the solar 

 diurnal variations observed at different stations on the earth's surface (even 



* Trans. Roy. Soc. Edin., vol. xxiv. p. 673. 



t See the projections for this lunation, Plate XLIIL, Trans. Roy. Soc. Edin., vol. xxii. 



| It is also partially due to the mode of combination — a lunation being considered in January, if 

 fifteen days were in that month, the other fourteen being in February or December, for which months 

 the range is considerably less than in January, and the maxima and minima occur at different times. 



