MAGNETIC OBSERVATIONS AT THE KEW OBSEEVATOEY. 447 



At Hobarton the results appear to be quite as consistent and systematic as those at 

 Kew or at Philadelphia, whilst in respect to the period of the year when the amplitudes 

 of the deflections are greatest, they present the phenomena in the reversed order, 

 October to March taking the position which at Kew and Philadelphia characterizes the 

 April to September half year. If we take the hours to 5, 12 to 17 as those of + 

 deflections, and 6 to 11, 18 to 23 as those of — deflections for both the half years at 

 Hobarton, we have the sums (disregarding their signs) of the semiannual monthly means 

 for October to March 4' '29, and for April to September l'-27 ; or if 23 to 4, 5 to 10, 11 to 

 16, 17 to 22 be taken, as possibly a preferable arrangement for the half year October to 

 March (retaining as before to 5, 6 to 11, 12 to 17, and 18 to 23 for the months April 

 to September), we have the sum 5''05 for October to March against l'"27 for April to 

 September. Under either arrangement the disparity is considerable, and appears deci- 

 sive in favour of regarding October to March as the half year of greatest deflection in 

 the lunar-diurnal variation at Hobarton. October to March is also the season of the 

 earlier occurrence of the maxima and minima : in both the presumed half-yearly charac- 

 teristics, therefore, the Declination at Hobarton, in the southern hemisphere, presents the 

 opposite features to the Declination at Kew in the northern hemisphere. 



I have added in Tables VII., VIII., IX. and X., at the close of this communication, 

 the annual and semiannual means of the lunar-diurnal variation of the Inclination and 

 of the Total Force at Kew and at Hobarton, although the number of years from which 

 these have been derived can hardly be considered sufficient to afford any secure founda- 

 tion for more than very general inferences in these two elements. For more assured 

 evidence we must await the continuation of the Kew Observations for the northern 

 hemisphere, and the magnetic investigations about to be commenced at Melbourne for 

 the southern hemisphere. 



In the meantime what is most worthy of our consideration is, that all the tables (in 

 the separate half years as well as in the whole year, and in all elements and all localities) 

 concur in the manifestation of that which is the great and distinctive characteristic of 

 the lunar-diurnal variation, viz. a double fluctuation in the twenty-four lunar hours in 

 equal or nearly equal divisions. It is this feature which must chiefly press itself on 

 the attention of those who would address themselves to the solution of the somewhat 

 difficult problem of the true theory of the moon's influence on the magnetism of the 

 earth. May it not be possible that this peculiar feature may be connected with the 

 duplex system of the terrestrial magnetism ? 



