II .
" On the Luinar iDiurnal Var-iation of Magnetic Declination at the Magnetic Equator .
" By JOHN ALLAN BRtoUN , F.R.S. , Director of the Trevandrum Observatory .
Ieceived Mairch 28 , 1860 .
This variationi , first obtained by M. Kreil , niext by myself , and afterwards by General Sabine , presents several anomalies which require careful considerationi , and especially a careful examination of the mnethods employed to obtain the results .
The law obtained seems to vary fromii place to place even in the same hemisphere and in the same latitude , and this to such an extent , that , for example , when the moon is on the inferior meridian at Toronto it produces a minimum of westerly declination ; while for the moon on the inferior meridiani of Prague and Makerstoun in Scotland it produces a maximum of westerly declination .
No two places have as vet given exactly the same result ; thoughi the result for each place has been confirmed by the discussioni of different periods .
Jn order to obtain the lunar diurnal action , it has been usual to consider the magnetic declination at any time as depending on the sun 's and moon 's hour-angles and on irregular causes .
Thus , if at conjunction , 110 be the variation due to the sun on the meridian , and ho be that due to the moon on the meridian , H , the variation for the sun at Ih , h1 for the moon on the meridiani of Ih , and so on ; it is supposed that we may represent the variations for a series of days by the following expressions , where the nearest values of A to the whole hour-angles are given:1st day .
E'o +h ' , +x'0 H1 ' +h ' , +x',1 H 123 + ?
h'22 +X'23 2nd day .
I1Ih/ ?
" 1"+"+ " It .
2n iy 11 0+h 123 + XIG o"iiI +A to +xI@ *e 23 +A " 21 XS 23 nth day .
IT ' +h " ?
xn H n ?
h~ + x ' .
11~ ?
h,6 +.IV , 010121 23 23 23 ) where x is due to irregular causes , and n is the number of days in a lunation niearly .
Summing these quantities we have approximately , § 23 0+ H1 +z232h+3x1 29 H23 ?
23h ?
ze ( A. ) and the means are , H+ a + ?
J 1XH+ - ] -,. .
H23+(I'23 ... .
( B. ) n nt n Here the hourly means are affected by the constant due to the total action of the moon on all the ineridians , and by variables depending on disturbing causes .
If , on the other hand , we arrange the series as follows , -I"o + A"0 -I tlo , H"3 03 4+ All + XI , , 1 " In this case 0 is the meau of tI observationis , of which 24 give the true means for the total solar influence , and the remaining n -25 being equally distributed through the hour-angles also give the mean approximately .
Instead , however , of comibining the observations in this way , the following method has been preferred .
Let , in the quantities ( B ) , ( 11 " ) _= HX +a+ 1-e ?
n+ Tnl x-i =+n __ I+ anso on it wll be -ob + Pre o+ inths ( tio)= there + ( areto)adssm 12.3 4-A + 23 _ ( H 23 ) 112 : + ( X2 =-1)_d z Stiomns ; ng the last two coliumns , we have =h + 23S(X ) Similarly we obtaiol ta tq +u te __ and so ons .
I. n-1 It will be observed that in these surmmations there are two assiimYptions ; one , that the lunar diurnal law is constant throughout the Ilunation , or series of lunations , for which the llneans are obtainled ; or that the quantity a in the expressions ( B ) is colnstanlt .
If this be not exact , then the quantity will contain the variation due to this n cause , and depend in part on the lunar hour-angle ; so that the mean ( H-I ) which is emnployed in taking the differences will eliminate part of the lunar action and partially distort the law .
The other assunmption is that the mean solar diurnal variation , represented by ( HOi ) , ( H , ) ... is nearly constant throughout the period ; for , if not , the differences due to such changes might be sufficient to mask any lunar law , the latter havinag a small range compared with the former .
Also it should be remarked that the means hoI hz1 , &amp;c. are combined with the irregular effect 23(x ) This effect , as far as it is due to n-i1 disturbance , we know obeys a solar diurnal law ; and if independent of lunar action , a sufficiently large series of observations might suffice to eliminate it , as combining with and forming part of the regular solar diurnal variation .
If , however , the series is not very large and the irregular disturbance considerable compare(d with the variation sought , it may be desirable to omit or modify the marked irregularities .
As regards the first assumption referred to above , the results obtained hitherto seem to show the error to be small , and the only way to determine its amnoulnt will be to consider it zero in the first instance , and thereafter a more accurate calculus may be employed .
For the second assumption , it is certaini that the solar diurnal law varies considerably in some cases withini a lunation .
At the magnetic equator , for example , the law of magnetic declination is inverted within a few weeks near the equinoxes .
The attempt to correct the error due to considerable change in the solar diurnial variation by taking the meanls , as has beenl done , from slhorter periods than a luniation , is liable to the serious objection that the resulting hourly means are affected uinequally byT the lunuar actioni , so that the sums ( A ) take the form , 2Ho + Y , ?
/1 + t.x0 &gt; ) , 1+ &gt; 2/ h 1 ... ... 2-12,3 + , ,2/ ?.x23 , where the second term in each expression is a variable .
In the discuissioni to which I am about to allude , the following plan has been followed .
The hourly means for the following series of weeks were takeni , naamelyml from Ist , 2nid , 3rd , and 4th weeks of the year .
2122 ) 2nd , 3rd , 4th , and 5th 35 ml , 3rd , 4th , 5th , and 6th The meanis of mii and l22 were them taken as normals for the 3rd or middle week , of 52n and m3 as ntornals for the 4th week , aid so on : these means were then employed for the differences from the corresponding hourly observations of the weeks to which they belonged .
With reference to the irregular effect , it is evidently desirable that we should know in the first instanUce whether it may not be a function of the lunar , as well as of ' the solar , lour-angle ; for this end it is essential in the first instanice to obtain the result including all the supposed irregular actions , and afterwards to eliminate these in the best manner possible .
In the discussion of the Makerstoun Observations I had substituted for certain observations , which gave differences from the mean beyond a fixed limit , values derived by interpolation from preceding and succeeding observations .
General Sabinie in his discussions has rejected wholly the observations which exceeded the limit chosen by him .
The omission of observations accidentally or intentionally , and the taking of means without any attempt to supply the omitted observations by approximate values , require consideration .
Let m be the true hourly mean for an hour A , derived from the complete series of n observations ; let ml be the mean derived from nI observationis , one observation o being accidentally lost ; then n-i nmt-o nf n-1 If , however , we supply the omitted observation by an interpolation between the preceding and succeeding observations , and if the interpolated value be o+x , we have , =nm+x mm + , ftx n The comparative errors of ml and mlr are therefore o-m x ; --and .
We may for any given class of observation determine the mean values of these errors .
Example:-At Hobarton , in July 1846 , the mean barometer for 3h ( Hobarton meanr time ) was 29 848 in .
, and the mean difference of an observation at that hour from the mean for the hour was 0403 in .
; if an observation had been omitted with such a difference , or for which o-m= 0 403 in .
, we should have an error in the resulting mean of ?
2 3=0 016 in .
, and the error might have been twice as great had the observation with the greatest difference been rejected .
If we now seek the error of mf , where the observation is interpolated , we shall find for the same month that the meani value of =0O005 in .
nearly ; whence the error 00002 in .
only , and the error would never exceed 0 001 in .
A similar though less advantageous resuLlt will be fou-nd in all classes of hourly observations .
In the case where observations are rejected which differ from the mean for the corresponding houir more than a giveni quantity , let us suppose , to simplify the qulestion , that the sums of n-I out of n observations for each of two suiecessive hours are each equal M , and that the observations for the same hours of the nth day are respectively mr+n and in+lt , where ml= M1 , 1 is the limit n1 beyond which observationis are rejected , and x is the excess of the observation to be omitted .
The mneans retaining all the observations re , M ?
l ?
i / 2i"M ?
m ' ?
+1-X iM+x+ , nn but if we reject the observation mn ' ++ x , we have =n , , in , .
nt-i It is assumed that nz1j9 12 0 ( any other hvpothesis of variationi would give the slame final result ) , and therefore the error of the change from the first hour to the second , when all the observations are retained , is n ; but if the observationi be rejected , the change is x mf+ 1Ia_ UMU This error , therefore , will be greater than the other if 1 &gt; tc ; so that the error in the resulting change from one hour to the next will be less by retaining an observationthan by rejecting it , if the difference from the preceding observation be not greater than the difference fIroyn the hourly mean ; that this will most frequently be the case will be obvious from the following fact:-At Makerstoun , in 1844 , at 1 A.M. the number of observations which exceeded the monithly means by 3 ' and less than double that , or 6 ' , was 99 , while the whole number which exceeded by more than 6 ' was only 16 .
It will be evident also that the difference 1 of an observation from the vorrespodling ' bourly 1'ne:an mnyv not be .
( duie to irregular causes , or to causes which affect the changes from one hour to the n-ext in a perceptible manner , but to gradual and regular daily change .
If we examine the daily means most free from irregnlar or intermittent disturbance , we shall find that they vary plus or miinus of the monthly meanr ; if the difference amounts to I in any case , then the whole observations of the day may be rejected though they follow the normal law .
By taking a proper value of I this case may not happen frequently , but cases like the following will .
At Jiobarton the daily means of magnetic declinationi differ in some months from the monthly means by 2 ' O nearly ; as the limit chosen by General Sabine is 2 ' 4 , any observation in such days differing by 01-4 from the niormal mean would be rejected .
The 25th and 26th days of March 1844 had beeni chosen by me as days free from magnetic disturbance , and following the normal law at Makerstoun ( Make .
Obs. 1844 , p. 339 ) , yet the means of horizontal force for these days differed 000064 and OOQ075 from the monthly means ; had the former quantity been the limit , all the observations on these days might have beenl rejected .
Altogether it appears to me that the method of rejecting observations beyond certain limits should not be employed at all , or if employed , only when interpolated observations are substituted ; and that this interpolation should constitute a seconid part of the discussion , the first incluiding all the observations- .
These considerations may appear somewhat elementary , but it is essential that results which present such anomalies as the lunar diurnal variation of magnetic declination should be obtained in a manner the most free from objection , even though the objections should touch on quantities of a second order compared with those obtained .
The discussion of which I now proceed to note the results , includes all the hourly observations without exception , made in the Trevandirum Observatory ( within a degree and a half of the magnetic equator ) during the five years 1854 to 1858 ; the second part of the discussion , in which days of great magnetic irregularity have been *I should note here my belief that a peculiarity noticed by General Sabine in his discussions as requiring explanation , namely , that the excursions of the declination needle east and west in the lunar diurnal variation have very different mangnitudes , is due to the rejection of ohservations , while the means by which the differences were obtained included the rejected quianitities .
' wholly rejected , not being conmpleted , I shall reserve the details for a more formal communication to the Royal Society .
The results obtainied are as follows : Ist .
At the maglnetic equiator the lunar diurnial law of magnetic declination varies with the moon 's declination and with the sun 's decliniation .
2nd .
This variation is so considerable that the attenmpt to combine all the observationis to form the mnealn law for the year gives results that are not true for any period .
Ihence evidently the impossibility of relating the laws at different places .
The so-called nean law for the year at Trevandrum obta.ined for the mooni furthest north , on the equator going south , furthest south , and on the equator going north , consists of three maxima and three minima , -a result wholly false , excepting as an arithmletical operation dtue to combination of very different laws .
3rd .
The lunar diurnal law va-ries chiefly with the position of the Sun , the variation being comparatively small with the position of the moon .
4th .
At the magnetic equiator the range of the variations is markedly greatest in the moniths of January , February , Novemnber and December , or about perihelion .
The following results are derived after grouping the means for differenit positions of the mnooni in periods of six months , October to March , and April to September ; they are therefore , for the reason given in the 2nd conclusion , not quite accurate ; but the change of the law from monith to mouth will be followed when the details are presented to the Society .
The following will give a general idea of the changes.5th .
TWhen the moon is furthest north .
a. About perihelion .
The lunar diurnal law of magnetic decliniation consists of two maxima when the moon is near the upper and lower meridians , the maximum for the latter being much the greatest ; of the two minima at inte-rmediate epochs , that for the setting mioon is the most marked .
6 .
Abotut aphelion .
The law consists of two nearly equtal minima iear the upper and lower transits : of the two intermediate maxima , that near the moonset is the most marked .
c. Thus the law about the winiter solstice is inverted about the * The decliniation is easterly at Tirevauiuntilll , auid the maixima indicate gr-eater e(tstebr/ ( leclination , suminmer solstice , and the one law passes into the other at the epochs of the equinoxes , exactly asfor the solar diurnal variation .
6th .
For the moon on the equator going south .
a. About perihelion .
The lunar diurnal law consists of two nearly equal maxima near the superior and inferior transits : of the two intermediate minima , the moonset minimum is by far the most mnrked .
b. About aphelion .
The law consists of two nearly equal minima near the superior and inferior transits : of the two internmediate maxima , that near moonrise is by far the most marked .
c. In this case also the laws for the solstices are the opposite of each other , and the one law passes into the other near the epochs of the equinoxes .
7th .
For the moon furthest south .
a. About perihelion .
The lunar diurnal law consists of maxima near the upper and lower transits , that at the upper transit being by far the most marked : of the intermediate minima , that near moonset is the greater .
6 .
About aphelion .
The law consists of two minima , the most marked at the inferior transit , the other about three hours before the superior transit ; and of two equal maxima , one near moonrise , the other near the superior transit , but varying little till 3 hours before the inferior passage .
c. In this instance the inversion is not so complete as in the other cases ; this , it is believed , will be found to be due to the fact that the change from onie law to the other takes place after the vernal and before the autumnal equinox ; so that in the means for six months , from which the aibove conclusions are drawn , the lunations following the law a are combined with those belonging to b. 8th .
The moon on the equator going north .
a. About perihelion .
The lunar diurnal law consists of two nearly equal maxima when the moon is near the superior and inferior meridians ; of the two intermediate minima , that near moonrise is by far the most marked .
6 .
About aphelion .
The law consists of two minima at the inferior and superior transits ; and of two maxima , the greatest at moonset , the other between the meridians of 16 6and 2 lIh ; betweeni these points there is an inflexion constituting a slight minimum .
c. In this case also the opposition of the laws is sufficiently well marked ; the only divergence from opposition being that due to the minior minimum about the meridian of 19h , due , it is believed , as nioted 7th c , to tbh partial combination of opposite laws in the aphelion half-year .
9th .
It will be observed that the variations of the law with reference to the moon 's declination or any given period of the year , consists chiefly in the differenice of the relative values of the maxima and ininima , the differences of epochs being small .
Thus for perihelion , the mooni fuLrthest nortli , the principal maximum occurs at the infe .
oior passage ; the moon on the equator going south , the two maxima are nearly equal ; the moon furthest south , the maximum at the superior passage is by far the greatest : on the equator going north , the two maxima are again nearly equal ; and so on for other epochs .
10th .
The moon 's actioni is chiefly , if not wholly , dependent on the position of the suni , or ( which is the same thing ) on the position of the earth relativJely to the sun ; and the law of the lunar action at the magnetic equator resemnbles in some points that for the solar action at the same epochs .
Thus about aphelion there is a minimum of easterly ( maximum of westerly ) declination produced by the lunar action , as well as by the solar action , for these two bodies near the suiperior meridian ; whereas abouLt perihelion both actions for the sun and mooni near the superior meridian produce maxima of easterly declination .
A like analogy holds for near the epochs of sulrise and moonrise .

