22 REPORT — 1860. 



Yearly Mean. March. October. 



1854 2-244 0-569 1/336 



1855 2-045 0-623 1-050 



1856 2-009 0-849 1*075 



1857 2*150 0-948 0989 



1858 2*414 1-265 1-091 



It will be perceived from this Table that the range for the month of March has 

 gone on increasing from 1854 till 1858, the range for the latter year being more than 

 double that for the former ; that while the minimum for the whole year occurred in 

 1856, that for March occurred in 1854, or before that year. In the case of the 

 month of October, the ranges differ little, that for 1854 being the greatest, and that 

 for 1857 the least. It is conceived by the author that this curious variation in 

 March and October is connected with a shift in the epoch of minimum diurnal 

 variation. If this epoch happen near the middle of the month, the range for the 

 mean of the month will be least; if it happen earlier or later, the greater range of the 

 preceding or succeeding periods will preponderate in the monthly mean. Should this 

 be the cause of the variation of range for March and October, it would follow that 

 the two superposed variations which produce the total variation may change their 

 relative values from year to year for a given place and for a given position of the sun. 



Lunar-diurnal Variation. — This variation, which was first remarked by M. Kreil, 

 and afterwards, though quite independently, by the author, has since been discussed 

 by General Sabine. The latter gentleman has made it a subject of inquiry, first, 

 whether the lunar-diurnal variation within the tropics obeyed different laws for the 

 moon north and south of the equator, like the solar-diurnal law ; and 2nd, whether 

 the decennial period could be perceived in the former as it is in the latter. His con- 

 clusions in both cases have been in the negative. Mr. Broun has discussed five 

 years' observations at Trevandrum, from which he arrives at the following results. 



1st. That the lunar-diurnal law varies with the moon's declination, but not to 

 the extent of inverting the law. In all cases there are two maxima of easterly de- 

 clination near the superior and inferior transits, and two minima for the moon near 

 the horizon. If we consider the period about the solstice of December, we shall find 

 that the greatest maximum occurs at the inferior transit for the moon furthest north, 

 and at the superior transit for the moon furthest south. The greatest minimum is 

 near moonrise for the moon on the equator going north, and near moonset for the 

 moon on the equator going south, while the minima are equal for the moon furthest 

 north and furthest south. The epochs also vary slightly. 



2nd. The lunar-diurnal law, which remains nearly constant as regards epochs for 

 all positions of the moon at any given season of the year, is the inverse in June of 

 what it is in December ; so that, for the sun farthest north, the lunar-diurnal law has 

 its maximum where for the sun furthest south it has its minima, the latter occurring 

 near the epochs of transit in June and July. In this way the lunar -diurnal law de- 

 fends on the position of the sun relatively to the ecliptic, and not (or little) on that of 

 the moon. 



The range of the lunar-diurnal variation is greatest near perihelion, which is just 

 the reverse of the solar-diurnal law ; this appears to depend on the moon's greater 

 proximity to the sun as the cause of its magnetic action. The range is least near 

 the epochs of the equinoxes, as for the solar-diurnal law. 



The cause of the great differences found by General Sabine in the laws for dif- 

 ferent places in the same hemisphere, is attributed by the author partly to the com- 

 bination of laws which vary considerably at the same place for different seasons. 

 The author also pointed out that General Sabine's failure to discover the decennial 

 period in the lunar-diurnal variation may be due to the fact that, before he com- 

 menced his discussion, he had first cut out all the disturbances beyond a certain 

 limit, so that a greater proportion were rejected in the years of greatest disturbance. 

 The decennial law is one affecting the regular diurnal variations, chiefly, through the 

 disturbance ; so that if the latter be omitted the effect should not appear (or appear 

 but slightly) in the former. The projected observations were exhibited to the Section. 



