ON BRITISH COLONIAL MAGNETIC OBSERVATORIES. 357 
required that no conclusions should be stated but such as were, at the 
least, tolerably assured by concurrent evidence drawn from the different 
quarters from which corroboration might be sought; and this implied not 
alone the completion of processes of reduction to which the observations from 
the different observatories must necessarily be submitted, but also the addi¬ 
tional labour of a coordination of the results obtained in different parts of the 
globe. The time, indeed, is yet Far distant at which anything like a complete 
view can be taken of what is capable of being elaborated from the immense 
mass of materials accumulated in the lust few years ; but it is possible that 
the time is arrived when a first and partial report might be made which may 
not be altogether unsatisfactory. 
1 have said that the request made to me at the Hull meeting was to prepare 
a report to be included in the volume containing the record of this year's pro¬ 
ceedings, and as is usual in such cases, to form the subject of a sectional com¬ 
munication. It has since beeu stated to me by the gentlemen whose office it 
has been to prepare the arrangements for this meeting, that a communication 
made to the Association on a subject on which it 1ms taken so strong an 
interest, would be preferably made to the members of the different Sections 
united at an evening meeting. Now it lias long been the cherished custom 
ul ( mr Association that its members should undertake cheerfully, and without 
hesitation, whatever duties nmy be allotted to them; and however I may rea- 
»"iiably distrust my own suitability for so prominent a position as that which 
jam placed in, I have been sensible that I could not with propriety decline it. 
t would, indeed, ill become your General Secretary to depart himself from 
a . u obligation which lie has so frequently urged on others; and assuredly 
here is no one who has had more experience than I have had, that the person 
wio does his best to discharge the duty assigned to him may feel sure of 
receiving from you the fullest and most considerate indulgence. 
n considering then how I might most profitably use the opportunity which 
ms een afforded me, I have thought that by confining myself to some one 
'ranch ol the inquiry, I might hope to be able to convey, in the time to which 
address ol this nature must necessarily he limited, a more satisfactory notion 
” i at ‘ n progress of accomplishment, than if I were to enter upon a more 
x ensive field; and I have selected as the topic most likely to interest the 
unbtTs of the Association generally, t he evidence which wc think we have 
mac • e . x ' s ^ ence a direct relation between the small periodical 
of^gnetic fluctuations which are obviously superimposed upon the mean state 
hid 6 i' 8 ma S net ' Hn, » an( f the two other bodies of the solar system, the 
ten-pqi' 1, i ni0011 ’ which, if magnetism he a cosmieal and not simply a 
relation 4 a ^ enc ^’ be expected to afford some indications of such a 
di,i°tT encil, S- ^erefore, with the Sun;—the first point to which I will 
variat'mn Ul p at i ent ' 0n 18 l ! 10 changewhich is found to take place in the diurnal 
i nor | hie needle in different months of the year. To make this the 
nation 0 11 1 ' 3 ’ P Br ^ a Pf» desirable that 1 should first give you a brief expla- 
g Ug p e , , c . rna * variation itself. It was long since observed that a magnet 
ofevervih 0rizo, !f :i ^- v .’ a compass needle for example, undergoes in the course 
four hnn*Jf 9 o Varia *> n * n ! ,s direction, returning at the end of the twenty- 
MJinetinif.- n , » P 09 * tl0n which it occupied at their beginning. This is 
hours of t| Ca i ' l0r<, fy variation, because the direction varies at the different 
cycle in J -* 1 , an< * sometimes diurnal variation, because its period, or the 
there is still' 01 i var ' a .*' on !S completed, is that of a solar day. Although 
fois perindiiUfl ta at is obscure in the mode in which the sun may produce 
ai fluctuation in the magnetic direction, we are at no loss to re- 
