]\gt ; INDEX SLIP . .
Charlee .
Analysis of the Results from the Kagnorographs Quiet\ldquo ; during the Twelve Years 1891 to 1902 .
Phil. Trnu , vol. , pp. ) from Kew rlos .
Magnotio Data from Flnlouth and CIzu , , vol. 204 1906 , pp. 37$-406 .
" " Quiot\ldquo ; tmouth o with Kow .
Phil. , vol. , 1905 , pp. S7$-4oe .
otor D , for lmoath Kew .
, Charleo .
Phil. Tran , vol. 204 , , pp. S7$-6 .
[ 373 ] X. ofth the Fahnonfh on ' Qniet\ldquo ; Day.s th , Yecrrs 1891 to 1902 .
CHARLES , Sc. D. , .
D. , ( From the National Physical ) Received November 23 , \mdash ; Read December 15 , 1904 .
CONTENTS .
Sections Pagc 1 .
Introduction . . . . . .
373 2 .
Secular change . . . . .
375 3-5 .
Non-cyclic effect . . . . . .
Solar diurnal inequality . . . . .
384 11 .
Vector diagrams . . . . .
387 12-14 .
Analysis of the diurnal inequality in Fourier series 393 15-16 .
Variation throughout the year ( Fourier series ) 400 17-20 .
Relations with meteorologic phenomena . . . .
402 S1 .
IN May , 1903 , I communicated to the Society a discussion of the results obtained from the Kew magnetographs during the eleven years 1890 to 1900 , on the five quiet days a month selected by the Astronomer Royal .
That paper will for shortness be referred to as ( A ) .
When dealing with data om a single station , even when representing many years ' observations , one is confronted by the possibility that the results may be appreciably modified by defects peculiar to the apparatus or to the methods of observation employed , or that they may be largely dependent on local conditions , and so of very limited applicability .
For these reasons alone , it was desirable to discuss on parallel lines data from a second English observatory .
Another reason for such an investigation was the disturbed magnetic conditions introduced at Kew in 1901 by electric trams .
In consequence of their action , there is unlikely to be any adequate opportunity for comparing simultaneous undisturbed data from Kew and the new magnetic observatory now under construction in Eskdalemuir .
Thus the most promising way of securing continuity is a minute comparison of Kew with some observatory now existent which is likely to continue free from tram disturbances for some time .
Both sets of considerations pointed to a discussion of the results obtained at Falmouth Observatory .
Magnetographs have been in constant operation at ( 381 .
) 4.5.06 Falmouth since 1891 .
They have been under the direct supervision of Mr. KITTO , the Superintendent of the Observatory , who has acted throughout as magnetic obser .
The behaviour of the vertical-force magnetograph has been indifferent , and the curves from it have not been measured ; but the declination and horizontal-force netographs have worked satisfactorily , and the data obtained from them on the Astronomer Royal 's quiet days have appeared in the Annual Reports of the Royal ornwall Pytechnic Society , ' and also in the Royal 's Proceedings , ' or the ' Reports of the National Physical Laboratory .
' The proposed investigation met with the hearty support of Mr. WILSON Fox , Honorary Secretary of the Royal ornwall Polytechnic Society , and of Mr. KITTO , and the latter kindly took a number of special curve measurements necessary the inquiry .
The Falmouth magnetographs , though resembling those at in the size of the magnets and in their general design , are by another maker , and differ in a variety of minor points .
Thus the scale of the declination magnetograms is considerably less open than at , and the temperature coefficient of the horizontal-force magnet so small that its sign even is uncertain .
Also the Falmouth magnetic chamber has a considerably smaller diurnal variation of temperature than that at .
Again , the position and climate of the two observatories are widely different .
is in the level plain of the Thames Valley , some 60 miles from the sea , but the magnetic chamber is only some 20 feet above mean sea level .
Falmouth Observatory , though close to the coast , is about feet above sea level .
possesses , for England , an almost continental climate , whilst Falmouth enjoys the equable temperature of the ornish coast .
From these considerations there is strong reason to believe that any phsnomenon observed in the records of both observatories is not due to instrumental defects or to purely local peculiarities .
Falmouth is situated in N. latitude , W. longitude .
Local noon is thus 20 minutes 18 seconds after Greenwich , 19 minutes 3 seconds after Kew .
An 11-year period , 1890 to 1900 , was employed in ( A ) .
For Falmouth I have employed a 12-year period , 1891 to 1902 .
This includes ] ] the data published when the investigation commenced , and secures a sun-spot minimum period , 1899 to 1902 , equal in length to that of sun-spot maximum , 1892 to 1895 .
The mean sun-spot frequency after WoLFER for the whole period and the above sub-divisions was as follows : \mdash ; , 1905.\mdash ; The working of the instrument was improwed in 1902 .
The curves for 1903 and 1904 been tabulated since the above was written .
The results appear satisfactory and are being published .
] GRAPHS ON ' ' QUIET ) ) DAYS DURIN TWELVE YEARS 1891 TO 1902 .
Secular Change .
S2 .
As in ( A ) , use will frequently be made of the letters and to denote respectively Declination , Horizontal Force , Northerly Component and Westerly Component of Force .
Only two of these elements , of course , are independent , and they are connected by the relation .
The system of absolute observations at Falmouth and the method of standardizin the curves are the same as at Kew ( see ( A ) , S 4 ) .
Table I. gives the mean yearly values of and as deducible from the diurnal inequality data in the annual Falmouth Reports , with the values of and calculated therefrom , and the values thence deduced for the annual changes .
The TABLE I.\mdash ; Secular Change .
absolute values of and expressed in C.G.S. measure ; the annual changes have for unit C.G.S. In the case of and the renccs from the corresponding annual changes at Kew are given in brackets , the sign indicating that the Falmouth change was numerically the larger .
The total decrease of declination between 1891 and 1900 al almouth was absolutely identical with that at Kew ; even in individual years the results at the two stations show a very fair agreement , the mean difference in.espective of sign being only .
As at Kew , there is subsequent to lS95 a marked slackening in the 376 DR. C. CHREE : ANALySIS OF RESULTS FROM THE FALMOUTH MAGNETOrate of secular change in D. The annual increase in at Falmouth from 1891 to 1900 was on the average about 10 per cent. larger than at Kew ; but in individual years the two stations show somewhat widely different results .
How much of this apparent difference is real it is difficult to say ; part is probably due to instrumental or observational uncertainties .
The decrease in at Falmouth from 1891 to 1900 was about 7 per cent. less , the increase in about 10 per cent. greater than at Kew .
The total changes from 1891 to 1902 are given by S , S The force of which these are westerly and northerly components lies in a vertical plane which is inclined to the geographical meridian at an angle of , and is directed north-easterly .
For the period 1891 to 1900 the angle becomes , or less than the corresponding angle for Kew ( see ( A ) , ) .
Non-cyclic Effect .
S3 .
As explained in ( A ) , S13 , secular change and annual inequality , if they exist , imply a progressive change in the value of an element from midnight to midnight of the average day .
There may also be apparent changes due to instrumental causes .
The contributions to the a-periodic daily variation at Falmouth from secular change and annual inequality are of the same small order of magnitude as the corresponding quantities at Kew ( ( A ) , S 18 ) .
The results of instrumental change are less exactly known than at Kew , but appear to be similarly small .
In the contributions from these causes constitute but a very trifling part of the observed non-cyclic effect on quiet days .
In the contribution from the secular change\mdash ; which tends , however , to reduce the observed effect\mdash ; is not relatively negligible .
I have , however , as in ( A ) , made no attempt to separate that part of the non-cyclic effect which is attributable to known causes .
Tables II .
and III .
give a complete analysis of the mean observed values of the non-cyclic effect on quiet days in and for each month and year of the 12-year period , while Table IV .
shows the corresponding mean values in , and for the whole year and the three seasons , winter ( November to February ) , summer ( May to August ) , and equinox .
When dealing in ( A ) with the non-cyclic effect in , I admitted that there was some room to doubt its bond fide character , but indicated several features which pointed to a real physical origin .
A comparison of Table II .
with Table V. of ( A ) tends to strengthen the latter view .
Whilst the sign of the mean non-cyclic effect fluctuates from month to month , and even from year to year , there is a clear tendency , as at Kew , for signs to predominate in winter and equinox , whilst in summer the mean non.cyclic effects at the two places are both negative and fairly similar in size .
GRAPHS ON ' ' QUIET\ldquo ; DAYS DURING THE TWELVE YEARS 1891 TO 1902 . .
HHdE VOL. CCIV.\mdash ; A. 3 DR. C. CHREE : ANALySIS OF RESULTS FROM THE FALMOUTH MAGNETO .
GRAPHS ON ' ' QUIET\ldquo ; DAYS DURING THE TWELVE YEARS 1891 TO 1902 .
For the whole year the mean value comes out for Falmouth as against for Kew .
If , however , we take the years 1891 to 1900 common to the two investigations , these figures become for Falmouth and for Kew .
TABLE Iy.\mdash ; Non-cyclic Effect .
Seasonal Values .
Means from 12 Years .
( Unit in case of force .
) Year .
Taking the whole 120 months common to the two investigations , the signs of the non-cyclic effects in at Kew and Falmouth agreed in 69 cases , and differed in only 22 , whilst in 29 cases the effect vanished at one or both of the stations .
Even if the whole of the observed non-cyclic effect were in every case ascribable to natural agencies\mdash ; which cannot be claimed\mdash ; agreement in could not be expected to occur invariably at stations so far apart as Kew and Falmouth .
S4 .
In the case of a comparison of Table III .
with Table VII .
of ( A ) shows an exceedingly close agreement .
The mean non-cyclic effect throughout the year at Falmouth is , it is true , some 10 per cent. less than that at Kew ; but when we take the common period , 1891 to 1900 , this difference is reduced to about 5 per cent. The tendency in the non-cyclic effect to be large in years of sun-spot maximum , and small in years of sun-spot minimum , and to be relatively low in June , July , August and December , is as clearly shown at Falmouth as at Kew .
Out of the whole 120 months common to the two investigations there are only 5 in which the non-cyclic effects at the two stations differ in sign , as against 99 in which there is agreement , and 16 in which the effect vanishes at one or both stations .
From 1891 to 1900 there were only 7 months when the non-cyclic effect was negative at Kew , and in 4 of these its amplitude was only , while in no case did it exceed ; yet out of these 7 months no less than 5 show a negative non-cyclic effect at Falmouth , the remaining two having zero values .
Thus even the incidence of the few exceptions to the general rule exhibits a remarkable agreement at the two stations .
In the case of the mean seasonal non-cyclic effects in Table IV .
the largest value appears in each element in the equinox , the smallest or most negative value in summer .
The latter phenomenon was also observed at Kew Table X but the maximum values in and occurred there in winter , and whilst the equinoctial value for and exceeded the winter value , the excess in the case of is distinctly less than at Falmoutk TABLE V.\mdash ; Diurnal Inequality TABLE VI.\mdash ; Diurnal Inequality of HorizontalGRAPHS ON ' ' QUIET\ldquo ; DAYS DURING THE TWELVE YEARS 1891 TO 1902 .
381 of Declination .
( West .
) Force .
Unit C.G.S. TABLE VIII.\mdash ; Diurnal Inequanequ of Westerw GRAPHS ON ' ' QUIET\ldquo ; DAYS DURING THE TWELVE YEARS 1891 TO 1902 .
Component .
Unit C.G.S. Component .
Unit C.G.S. S5 .
It is natural to suppose\mdash ; as was in fact remarked by myself when first discussing the phenomenon*\mdash ; that the non-cyclic increase in on quiet days is connected in some way with the considerable decrements in , which in the majority of instances accompany magnetic storms .
It would appear , however , that the mean size of the non-cyclic effect on the Astronomer Royal 's quiet days for a cannot be inferred from the character of that year as disturbed or otherwise .
For example , 1894 was much more disturbed than the years immediately preceding or following it ; but the mean non-cyclic effect in at both Kew and Falmouth is distinctly less for 1894 than for 1893 or 1895 .
There were , in fact , only 4 months out of the 12 in which the non-cyclic effect in 1894 exceeded that in 1895 , and only 2 months in which it exceeded that in 1893 .
Again , during the four sun-spot minimum years , 1899 to 1902 , few months showed anything that would be classified as a magnetic storm in years of sun-spot maximum , and yet the proportion of months showing a decidedly positive non-cyclic effect is the same as for the disturbed year 1894 .
Solar Diurnal Inequality .
S6 .
The diurnal inequality figures in the annual Falmouth Reports include the non-cyclic effect .
This has been eliminated in preparing Tables V. and VL in the way described in ( A ) S18 .
The hours , it is important to notice , are Greenwich mean time .
Values of and are not given in the Falmouth Tables .
The hourly values in Tables VII .
and VIII .
were deduced from those in Tables V. and VI .
by means oi the relations\mdash ; the units being and 1'\mdash ; 0.947@H\mdash ; I.74ID , these being the special forms taken by the general formulae when and are assigned the mean values at the foot of Table I. As in ( A ) , and are counted positively to the west , and to avoid decimals the unit of force employed in the tables is .
In reality is measured only to and to , so the tables go a figure beyond the observations .
This is completely justified so far as Tables V. to VIII .
are concerned , because the figures assigned to any one month are means from 12 years , and the yearly means for the two sun-spot periods are based each on 48 months .
In addition to the mean hourly values , the tables give the range ( defined as the ' B. A. Report 1896 , ' p. 212 .
GRAPHS ON ' ' QUIET\ldquo ; DAYS DURING THE TWELVE YEARS 1891 TO 1902 .
diflerence between the greatest and least mean values ) and the numerical sum of the differences of the 24 hourly values from their arithmetic mean .
All pronounced maxima and minima values are printed in heavy type .
In most months two maxima and two minima are distinctly shown by each element , but sometimes only one pair appears .
The phenomena are so similar to those at Kew , that a brief discussion will suffice .
Declination .
S7 .
From October to March there are in Table V. , as at Kew , two pronounced maxima and mininna , and in one or two other months there is a suspicion of more than one maximum and minimum .
In the mean diurnal inequality for the year there is but a very short interval between the night maximum and minimum .
The time at which the needle reaches its extreme westerly position is nearly constant throughout the year .
The extreme easterly position is assumed in Decelnber about 10 , but in all the other months of the year about 8 to 9 .
The hours of maximum and minimum look later than at Kew , but that is mainly , if not wholly , due to the difference in local time .
In addition to the December minimum , prominent in all the elements , the ranges and the sum of the differences show , as at Kew , a second inconspicuous minimum at midsummer .
The interval between the two apparent maxima in the ranges is , as at Kew , longer than that between the two maxima in the sum of the differences , the latter appearing at both stations in June and August .
The monthly ranges and sums of the differences appear , on the whole , a trifle smaller than at Kew .
This is really attributable to the fact that the mean ) frequency for the Fallnouth period 1891 to 1902 was slightly less than that for the Kew period 1890 to 1900 .
During the common period 1892 to 1895 , in the mean diurnal inequality for the year , the Falmouth range was the greater by , whilst the Kew sum of the differences was the greater by .
As these differences represent only about 1 per cent. of the quantities themselves , the only safe conclusion is that the amplitudes of the diurnal inequality of declination at Kew and Falmouth are exceedingly nearly equal .
Horizontal Force .
S8 .
Two maxima and minima are shown from October to April , somewhat doubtfuly in April ; in the remaining five months there is at least a retardation the fall of about 3 or 4 .
The mean inequality for the year shows two maxima and minima , but the morning maximum and minimum are poorly defined .
The least value of the day occurs about 10 or , earlier in summer than in winter .
From November to February the largest value is met with at 6 or 7 but throughout the equinoctial and summer months the absolute maximum occurs VOL. CCIV.\mdash ; A. 3 at 7 or 8 .
The slight apparent tendency in the hours of maxima aid minima to be later than at Ksw is again ascribable to the difference in local time .
As at Kew , the range seems nearly stationary from May to August , sum of the differences from May to July .
In every month of the year the range and the sum of the differences , especially the latter , are larger than at Kew .
Judging by the common period 1892 to 1895 , the range in the mean diurnal inequality for the year at Falmouth is the greater by some 13 cent. Northerly and Westerly Components .
S9 .
The diurnal inequality in is similar to that in , but a second maximum and minimum in the morning are generally recognisable even in summer .
The chief minimum , near noon , and the afternoon maximum occur somewhat later than in H. The range and the sum of the differences show more clearly than in a secondary minimum at midsummer ; from October to March they are both decidedly larger than in , but at midsummer the difference is , if anything , the other way .
The general character of the diurnal inequality in resembles that in D. The afternoon maximum and the forenoon minimum\mdash ; the only pronounced turning points during the seven months , March to September\mdash ; seem slightly later than in D. The mean diurnal inequality for the year presents a very slow rate of change from 1 to 3 , but only one maximum and minimum are shown .
In the amplitude of the range there is at least a suggestion of two maxima in summer , with an intervening minimum , but the sum of the differences shows only a single maximum in June .
The range in is distinctly larger than in in the mean diurnal inequality for the year , and in every month except November , December and January , but the sum of the differences is larger in than in during 7 months of the 12 , and in the mean diurnal inequality for the year .
S10 .
For purposes of comparison of different elements at different stations , the annual variation in the ranges and in the sum of the differences is most conveniently exhibited by representing the 12 monthly values as percentages of their arithmetic mean .
This is done in Table IX .
The small figures in the and columns give the excess of the Falmouth percentages over the corresponding Kew percentages in Table XVIII .
of ( A ) .
The range and the sum of the differences vary pretty similarly , but the principal maximum is higher for the 24 differences than for the ranges , .
especially in W. There is not much difference between the different elements , but the midsummer values are decidedly higher and the midwinter values decidedly lower in than in , i.e. , the annual variation is greater for the former element than the latter .
The Falmouth percentages are , on the whole , distinctly above the Kew in winter and below .
them in summer , the difference being greater in than in D. This .
means that annual variation is less marked at Falmouth than at Kew .
GRAPHS ON ' ' QUIET\ldquo ; DAYS DURING TWELVE YEARS TO 1902 .
TABLE IX .
Monthly Relative Values , 1891 to 1902 ( mean from 12 months ) .
February August September .
October November .
Vector S 11 .
The mode of variation throughout the year in the diurnal inequality of the magnetic forces in the horizontal plane is conveniently illustrated by vector diagrams ( ( A ) , S29 ) .
The accompanying diagrams refer to the mean monthly diurnal inequalities for Kew ( 1890 to 1900 ) and Falmouth ( 1891 to 1902 ) , thick lines for the former , thin lines for the latter .
NS and EW represent respectively the geographical north-south and east-west directions .
Their intersection ( thick lines for Kew , thin for Falmouth ) answers to the origin .
The crosses and the numbers represent hours counted from midnight as the line from the origin to any number.or cross representing in magnitude and direction the disturbing force to which the diurnal inequality may be ascribed .
The line drawn from to the corresponding origin marks the direction of the magnetic meridian .
For the periods considered the mean values of the declination at Kew and Falmouth were respectively and .
The distances of the points marked from their co1responding origin represent in all cases .
Except in the case of December and January the distance between the origins for and Falmouth would represent The hours are G.M.T. , so that 12 , for instance , on a Falmouth curve answers to 11h . .
local mean time , but on a Kew curve to 11h . .
local mean time .
If local time had been used , corresponding Kew and Falmouth radii would have shown an even closer approach to parallelism than they do .
a FEBRUARY MARCH GRAPHS ON ' ' QUUIE\ldquo ; DAYS DURING THE TWELVE TO 1902 Ivl 390 .
C. CHREE : ANALySIS OF RESULTS FROM THE FALMOUTH GRAPHS ON ' ' QUIIE\ldquo ; DAYS DURDUG GHE TWELVE YEARS 1891 TO 1901 .
392 DR. C. CHREE : ANALYSIS OF RfflSULTS FROM THE FALMOUTH MAGNETONOVEMBER le DECEMBER GRAPHS ON ' ' QUIET\ldquo ; DAYS DURING THE TWELVE YEARS 1891 TO 1902 .
Perhaps the most comprehensive way of describing the conspicuous seasonal variation in the part of the curve which answers to the night hours is by reference to the two peaks that tend to form , one .
about 6 , the other about 7 .
From November to February the former is dominant .
In March the evening peak begins to assert itself , and its predominance remains undisputed during the summer months , whilst the morning peak tends to disappear .
In October the morning peak becomes again apparent .
In most months there is at least an indication of a subsidiary peak near midnight .
The curves are simplest and most symmetrical in June and July .
In August the afternoon peak is remarkably sharp both at Falmouth and Kew .
In October a very small loop appears , a retrograde movement setting in about midnight .
The loop enlarges in November and becomes most prominent in December , contracting in January and February .
In March there is a very tiny loop in the Kew curve shortly after midnight , and a curious indentation .
The Falmouth curve shows only the indentation .
On the other hand , the April curve shows a tiny loop about 7 at Falmouth , whilst at Kew there is only a peak .
The Kew curves for three months , December , March and June , appeared in ( A ) .
Vector diagrams for Greenwich were given by Sir G. B. AIRy for each month of the year , for the periods 1841 to 1847 and 1848 to 1857 , treated separately , latert for the four months January , April , July and Octpber in each of three successive years .
In making any comparison it should be noticed that AIRY counted time from Gottingen noon in the earlier curves , and from Greenwich noon in the later .
The different Greenwich curves for the same month of the year differ somewhat widely amongst themselves , and none of them resemble very closely the corrending Kew or Falmouth curves .
There is , on the other hand , a pretty close resemblance between individual Kew and Falmouth vector diagrams and the corresponding ones for Parc St. Maur , from the period 1883 to 1897 , given on p. 261 of Professor MASCART'S ' Magnetisme Terrestre .
' Analysis of the Inequality in Fourier Series .
S12 .
As in ( A ) , S31 , the analysis is supposed to be effected in the two alternative forms where represents Greenwich mean time counted from midnight , and an hour is equivalent to .
If local time were used , the corrections required to the phase angles would be obtained by adding to those given in ( A ) , Table XIX .
; for the addition would be , and so on .
'Phil .
Trans. ' for 1863 , Plates 18 and 19 .
'Phil .
Trans. ' for 1885 , Plate 74 .
VOL. CCIV.\mdash ; A. TABLE X.\mdash ; Coefficients and Angles in Fourier Series January February .
Septembel ' October November Winter Equinox .
TABLE XI.\mdash ; Coefficients and Angles in Fourier Series Expansion GRAPHS ON ' ' \ldquo ; DAYS DURING THE TWELVE YEARS 1891 TO 1902 .
Expansion of Diurnal Inequality of Declination ( 1891 to 1902 ) .
of Diurnal Inequality of Horizontal Force ( 1891 to 1902 ) .
Unit Period , Period , Tables X. and XI .
give the values of the coefficients answering to the diurnal inequalities of and for the individual months , the three seasons , and the entire year from the whole 12 years ' observations .
The values of the and constants were calculated first , the calculation going a figure further than appears in the tables , and the and constants were thence deduced .
The phase angles cannot claim an accuracy of 1 ' , especially the monthly values of and ; reasons for recording the minutes are given in ( A ) , S32 .
Table XII .
, giving Fourier coefficients for the diurnal inequalities in and , is confined to the seasonal and yearly data .
TABLE XII.\mdash ; Coefficients and Angles in Fourier Series Expansion of Diurnal Inequality of Northern and Western Components .
( Unit ) The difference between local time at Falmouth and Kew is a little in the way of comparison .
But careful study shows that many of the apparently irregular fluctuations in the values of and , which might be ascribed to observational uncertainties , appear to a greater or less extend at both places .
For instance , going to the nearest degree , we have the following consecutive values for in Whilst amongst the values of in we have : GRAPHS ON QUIET\ldquo ; DAYS DURING THE TWELVE YEARS 1891 TO 1902 .
The amplitude of the 6-hour term at midsummer being of the order in and in , I was originally disposed to regard irregularities such as the above at Kew as mainly , if not entirely , accidental .
It is now clear that whilst the irregularities may be accidental in the sense that they arise from special features in the days selected by the Astronomer Royal , they cannot be wholly , or even mainly , assigned to errors of measurement , or purely local causes .
A reference to Table XX of ( A ) shows that the amplitudes of the several terms in are closely alike at Kew and Falmouth .
In the case of a reference to Table XXII .
of ( A ) shows that the amplitudes of the 24-hour and 12-hour terms are decidedly larger at Falmouth than at Kew ; but the 8-hour and 6-hour terms have similar amplitudes at the two stations .
S13 .
Table XIII .
expresses the monthly values of the coefficients in and as percentages of their arithmetic means , and the three seasonal values as percentages TABLE XIII.\mdash ; Variation of Fourier Coefficients throughout the Year ( 1891 to 1902 ) .
January .
of their means .
The small figures give the excess of the Falmouth percentages over the corresponding Kew ones , Table XXIV Except in the case of , where there are apparent discontinuities , the irregularities in the difference between the Falmouth and Kew percentages are surprisingly small .
In for the differences hardly warrant a conclusion , but on the for being a partial exception\mdash ; the Falmouth values are above the Kew in winter and below them in summer .
Table XIV .
gives the ratios borne by , C3 and to in the mean diurnal inequality for the year , and also the excess of the ratios over the corresponding Kew TABLE XIV.\mdash ; Relations between Fourier Coefficients in mean Diurnal Inequality for the Year ( 1891 to 1902 ) .
ones .
In and the 24-hour term is relatively more important at Kew than at Falmouth , but the reverse is true in the case of H. The differences between the two places are however very small , especially in the case of and N. Table XV .
gives similar data for the seasonal diurnal inequalities .
Corresponding Kew data existed only for and H. The Kew-Falmouth differences have the same TABLE XV.\mdash ; Relations between Fourier Coefficients in Diurnal Inequalities for the Seasons ( 1891 to 1902 ) .
sign for an three seasons as for the year ; they are larger , however , in winter than in summer .
Except in winter the 12-hour and 8-hour terms are decidedly more important relatively to the 24-hour term in and than in and ; and the 8-hour and the 6-hour terms are relatively more important in than in N. In cases the 8-hour and 6-hour terms are relatively more important in winter than in summer .
S14 .
To facilitate comparison of the phase angles at Kew and Falmouth , GRAPHS ON ' ' QUIET\ldquo ; DAYS DURING THE TWELVE YEARS 1891 TO 1902 .
Table XVI .
gives the values of the angles for the mean diurnal inequalities of the three seasons and the year when Falmouth solar time is used ; it corresponds to Table XXVIII .
of ( A ) for Kew .
TABLE XVI.\mdash ; Seasonal Values of the Angles in Fourier Coefficient Expansions of Diurnal Inequality ( 1891 to 1902 ) when Falmouth Solar Time used .
Table XVII .
gives the local solar times of the earliest maximum in the day for the several terms .
These times , as in ( A ) , Table XXIX .
, given only to decimals of an hour , it being doubtful whether higher accuracy can be claimed for individual results , especially in the terms of shortest period .
In calculating , however , the differences between the local times of occurrence at Falmouth and Kew , I employed the complete TABLE XVII.\mdash ; Times of Occurrence of First Maximum in the Terms of the First Four Orders in Fourier Expressions for Diurnal Inequality for the Period 1891 to 1902 ( Falmouth Local Time ) .
hours .
angles in Table XVI .
and the corresponding Table in ( A ) , and , in converting the differences into time , went to the nearest minute .
These differences , in minutes , appear in small type in Table XVII .
, the sign denoting later occurrence at Falmouth .
The approach to coincidence in the local times of occurrence of the maxima at the two stations is truly remarkable .
The algebraical mean of all the * The time 1 1 .
assigned to the Kew 24-hour -term in winter is in error .
It should be differences is somewhat under minute , which answers to about millim. on the magnetograms .
As the sign of individual differences , owing to their exceeding smallness , must be largely fortuitous , the only safe conclusion is that whilst there is a preponderance of signs ( i.e. , of cases when the event is later at Falmouth than at Kew ) , the differences may be accidental and are certainly exceedingly small The above results refer to the average year , and no inference can safely be drawn as to the differences between Kew and Falmouth in different individual years .
Variation throughout the Year ( Fourier Series ) .
S 15 .
The ranges of the diurnal inequalities , the sums of the 24 differences from the mean , and the amplitudes of the 24- , 12- , 8- , and 6-hour terms have had their annual variations expressed in Fourier series , the constants being calculated ffom the monthly values .
Table XVIII .
gives the results for the 12-year period .
The time is counted from midnight of December 31 to January 1 , and a month is equivalent to The amplitudes of the annual and semi-annual terms\mdash ; which suffice to express the variation very completely\mdash ; are denoted respectively by and .
The table gives the ratio of to , and also of and to , the mean of the 12 monthly values of the element concerned .
The reservations necessary are discussed in ( A ) , S 37 .
The excess of the values of , \amp ; c. , at Falmouth , over those at Kew are given as usual in small type ; in the case of and these differences are omitted for as too uncertain .
Except in the annual term in the phase angles differ comparatively little amongst themselves ; the same phenomenon appeared at Kew , Table XXX Considering the smallness of , one would be disposed to suspect that the large difference between the phase angles of the annual term in that element and the other angles , though appearing in both and , was due to defects in the observations or reductions .
Exactly the same phenomenon appeared , however , in the corresponding case at Kew ( see ( A ) , Table XXX The ratios borne by the amplitudes of the annual and semi-annual terms to the mean value of the element at Kew and Falmouth differ comparatively little .
But if we exclude , where the element and its variable part are both very small , we see that on the whole the annual term at Falmouth is somewhat less important as compared either to the semi-annual term or to the absolute mean than it is at Kew .
This is what Tables IX .
and XIII .
would lead us to expect .
S16 .
The dates of occurrence of the maximum in the annual terms , and of the earliest maximum in the semi-annual ternoe , are given in Table XIX .
explained in ( A ) , S38 , these dates may be uncertain to the extent of a whole day , for they are based on a treatment which disregarded differences between the lengths of the several months .
GRAPHS ON ' ' QUIET\ldquo ; DAYS DURING THE TWELVE YEARS 1891 TO 1902 .
TABLE XIX.\mdash ; Annual Variation ( 1891 to 1902 ) .
Date of Occurrence of First Maximum .
Sum of 24 differences .
The figures in small print denote the differences in between the times of occurrence of corresponding maxima at Falmouth and Kew , the plus sign indicating a later occurrence at Falmouth .
The phase angles , though recorded in Table XVIII .
and in Table XXX .
of ( A ) only to the nearest whole degree , were realy worked out to minutes of arc , and the differences in Table XIX .
were derived from these more complete angles .
The differences are also nearly free from several sour of uncertainty which enter nearly equally into the absolute values of the phase angles at Kew and The algebraic mean of the differences , when taken to the nearest 1 ' and converted into time , represents hours , but if the somewhat outstanding result from the annual term for in be omitted , this becomes hour .
Thus no certain conclusions can be drawn as to whether the maxima in the annual variation appear earlier at Kew or at Falmouth .
The truly remarkable approach to coincidence between the Kew and Falmouth dates certainly points to the conclusion that the annual and semi-annual terms represent very real phenomella , determined by laws which vary but slightly over large areas .
These results , it must again be pointed out , relate to the mean year of a sun-spot cycle , and so throw no light on the differences that may exist between individual years ' results at the same or at different British stations .
with Meteorological Phenomena .
S17 .
Table .
gives the mean values of certain meteorological elements at Falmouth for the entire period 1891 to 1902 , the sun-spot maximum period 1892 to 1895 , and the sun-spot minimum period 1899 to 1902 .
It also gives the algebraic excess of the mean for the sun-spot minimum period over that for the sun-spot maximum period .
Corresponding excess data for Kew are given ( a ) for the same as at Falmouth , ( b ) when the un-sPot consists of the three GRAPHS ON ' ' QtIIET\ldquo ; DAYS DURING THE TWELyE YEARS 1891 TO 1902 .
40.3 TABLE .\mdash ; Mean Values of Meteol.ological Elements .
years 1890 , 1899 , and 1900 .
The last set of Kew results are deducible from Table XLVI .
of ( A ) .
The Falmouth data are from the tables in the annual ' Reports of the Royal Cornwall Polytechnic Society .
' They are deduced from the curves given by self-recording instruments of the same pattern as at Kew .
There are other temperature data in the Falmouth Reports ; but they relate to thermometers in a Stevenson screen and are less comparable with the Kew data .
The vapour-pressures and barometric heights in the table are in inches of mercury ; the temperatures are in degrees Fahrenheit .
The mean height of the barometer for the whole 12 years was higher than the mean height during either the sun-spot maximum or minimum periods , and the same phenomenon appeared at Kew whether the periods were those of case or case ( b ) .
Again , the mean vapour-pressure for the whole period did not differ from that for the sun-spot minimum period by more than inch either at Falmouth or Kew .
Thus in these two elements there is less evidence for a cormection with sun-spot frequency than might appear at first sight .
As regards the mean daily range of temperature , whilst all the differences are of one sign , they vary much in size , and the mean from the 11 years 1890 to 1900 at Kew was smaller than that from either sun-spot period in ( b ) .
S18 .
In the case of the mean annual temperature , the long-period mean was intermediate in all three instances between those for the periods of sun-spot maximum and minimum , and the differences between these two latter means are fairly similar .
The existence of a difference between the average mean temperature of years of sun-spot maximum and minimum is accepted as an established fact in HANN'S ' Climatology.'* According to HANN'S figures the mean excess of temperature in a period of four sun-spot minimum years over a period of four sun-spot maximum years , situated relativelv to maximum and minimum as in the case of the Falmouth of years , is for a tropical station no less than C. , the difference between the two * English Translation by WARD , p. 405 .
extreme years of an 11-year period amounting to C. At an average extratropical station , according to HANN , this latter figure is reduced to C. Allowing tentatively a corresponding reduction on the C. , it becomes C. Thus the differences recorded in Table .
are not merely of the normal sign , but even of somewhat the same order of magnitude as they should be according to HANN at an average extra-tropical station on the average of a number of sun-spot periods .
The fact that in Table .
a rise in the mean annual temperature is associated with a diminution in the mean daily range may be accidental , but is possibly worth fuller investigation .
S19 .
Table XXI .
shows the variation throughout the year at Falmouth in the mean vapour-pressure for the 24 hours , in the daily range of temperature , and in the amplitudes of the 24-hour and 12-hour terms in the diurnal inequality of temperature .
TABLE XXI.\mdash ; Monthly Relative Values ( mean from 12 months ) .
euahmsemrat 1 The mean monthly values are expressed as percentages of their arithmetic mean .
The vapour-pressure and temperature-range data are based on the tables in the Falmouth Reports for 1891 to 1902 .
The data for and relate to the period 1871 to 1882 , and are calculated from mean absolute monthly values given by General STRACHE The figures in the columns headed " " magnetics\ldquo ; are means from the data for and in Tables IX .
and XIII .
The small figures give the excess over corresponding percentages for Kew .
The magnetic data employed here for Kew refer , like the Falmouth , to and only , and so differ slightly from those given in Table XLVIII .
of ( A ) .
There is a general resemblance between the annual variations of vapour-pressure , temperature range , and magnetic range ; but , as at Kew , the ' Phil. Trans. ' for 1893 , p. 644 .
ORAPHS ON ' ' QUIE \ldquo ; DAYS DUliING TWELVE yEARS TO 1902 .
ter minimum occurs earlier and is more strongly marked in " " magnetics\ldquo ; than in either meteorological element .
The resemblance between the annual variations in the amplitude of the 24-hour term in temperature and magnetics was somewhat striking at Kew S68 .
This resemblance is also visible in Table XXI .
, but is decidedly less close .
In the case of the annual variations in temperature and magnetics are less conspicuously unlike than at Kew , but still wide apart .
Vapour-pressure and temperature range agree with the several magnetic quantities in showing at Falmouth higher percentages in winter and lower percentages in summer than at Kew .
The Falmouth amplitude in temperature shows relative to the Kew increased percentages at midwinter , but it differs entirely from the magnetic in showing still more enhanced at midsummer and largely diminished percentages at the equinoxes .
The mode of difference between the Kew and Falmouth annual variations in is diametrically opposite for temperature and magnetics .
S20 .
As the existence or non-existence of parallelism between the modes of daily variation of temperature and terrestrial magnetism is of tance in iuquiries as to how the diurnal magnetic inequality comes into existence , attention may usefully be called to some further points .
As remarked in $S 7 and 8 , the mean diurnal ranges of declination at Kew and Falmouth are almost exactly equal , whilst the diurnal range in is somewhat greatest at Falmouth .
The mean difference , however , between the daily maximum and minimum of temperature from 1891 to 1902 was only at Falmouth as against at Kew ; whilst taking the mean diurnal inequality for the year from the Meteorological Office 's ' ' Hourly Readings during the sun-spot maximum period 1892 to 1895 , the range was at Falmouth , at Kew .
Again , the amplitudes of the principal Fourier terms in the mean magnetic diurnal inequality at Falmouth are very similar to those at Kew in and larger on the whole in H. In the case of temperature , however , General STRACHEY obtained the following values for the amplitudes of the three principal terms during a common period 1871 to 1882:\mdash ; FalmouthKew 12-hour term .
Further , in S 14 we saw that the local times of occurrence of the maxima in the Fourier terms in the diurnal inequalities for and were as nearly as possible identical at Kew and Falmouth .
General STRACHEY , the other hand , also ' Phil. Trans .
cit. , p. 645 .
employing local time , found that the phase angles were in every month of the year larger at Kew than at Falmouth in the 24- and 12-hour temperature terms , whilst the Kew angle for the 8-hour term was the larger in eleven months of the year .
The mean of the 12 monthly differences between the Kew and Falmouth phase angles from General STRACHEY'S figures are as follows:\mdash ; It is thus clear that temperature phenomena near the earth 's surface in England are modified by local conditions , such as proximity to the sea coast , in a degree to which there is no parallel in the phenomena of terrestrial magnetism .
As in ( A ) , all the Fourier coefficients made use of in the paper were calculated directly from the formulae .
In the laborious arithmetical work involved in this and in some other parts of the paper I have received valuable assistance from Mr. B. FRANCIS , Assistant in the Observatory Department .
Amidst such a multiplicity of figures absolute freedom from error can hardly be hoped for , but all calculations have been gone through at least twice , and every reasonable precaution has been taken to avoid lnistakes .

