[ 187 ] V. Discussion of Kew Magnetic Data , especially the Diurnal Inequalities Horizontal Force and Vertical , from Ordinary Days of the Eleven Years 1890 to 1900 .
ByC .
Chree , Sc. D. , LL. D. , F.R.S. , Superintendent , Kew Observatory .
Received November 25 , 1915\#151 ; Read January 27 , 1916 .
Contents .
SS Page 1 .
Introduction ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .187 2 .
Description of material .
Days not used ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ..189 3 .
Methods of tabulation employed ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .192 4 .
Non-cyclic change ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .
194 5\#151 ; 11 .
Diurnal inequalities .
Tables and curves ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .197 12-14 .
" " Vector diagrams ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ..225 15-18 .
Annual inequality ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ..234 19-22 .
Fourier coefficients .
240 23,24 .
Sunspot relations .
Wolf 's formula ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .252 25 , 26 .
Daily range .
Tables ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... 256 27 .
" " Sunspot relations ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .262 28-30 .
" " Frequency of occurrence ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .267 S 1 .
In 1901 the development of electrical traction in West London rendered it clear that unless conditions altered in an unexpected direction , no further magnetic records from Kew Observatory would be sufficiently free from artificial disturbances to be an altogether satisfactory medium for the minute study of phenomena such as the regular diurnal variation .
The time seemed to have come for taking stock of the records obtained .
The measurement of magnetic curves formed no regular part of the work at Kew Observatory until 1890 .
Prior to that date the Annual Reports contained only a summary of the results of the absolute observations .
Since 1890 the programme of work has included the measurement of the magnetic curves for 5 " quiet " days a month .
Until a few years ago these days were selected at Greenwich by the Astronomer Royal .
Now they are selected under international auspices at De Bilt , the central station of the Meteorological Institute of the Netherlands .
Since Kew Observatory was transferred to the Meteorological Office , diurnal inequalities have been published for each month of the year for D ( declination ) and H ( horizontal force ) .
Previously the inequalities published were VOL. CCXVI .
A 542 .
2 D [ Published March 28 , 1916 .
confined to the year , winter ( October to March ) and summer ( April to September ) .
Vertical force ( V ) and inclination ( I ) suffer more than D and H from artificial electric currents , and no inequalities have been published for them since 19.01 .
It was decided to begin by analysing the measurements that had already been made of the " quiet " day curves for an eleven-year period , 1890 to 1900 .
The necessary labour was almost entirely arithmetical , and it was completed with the aid of the Observatory staff without extraneous financial assistance .
The results were embodied in a paper* published in 1903 .
It had gradually been recognised that diurnal variations derived from quiet days are not identical with those derived from all days , or from all days but those of large disturbance .
It became increasingly obvious that the Kew records would not be fully utilised until the study was extended to other than quiet days .
The magnitude of the task was not at first fully realised , and the original programme seems to have embraced all the accumulated data , limiting the enquiry however in the first instance to the declination .
At all events the list of Government Grants for 1903 to 1904 includes one of \#163 ; 82 10s .
" to work up declination ( magnetic ) results obtained at Kew from 1857 to 1900 .
" The work was practically confined to the 11 years 1890 to 1900 , and the grant was exhausted before it was completed .
The curves were divided into " ordinary " and " disturbed .
" An ordinary day was one in which the general trend of the diurnal variation was clearly recognisable , so that when the trace was oscillatory it could be fairly replaced by a freehand pencil curve of moderate curvature .
It had been the practice , when sensible oscillations occurred on a selected quiet day , to smooth the curve , replacing it by a pencil trace , so the procedure adopted with the ordinary day curves was no innovation .
Smoothing was , however , done much more extensively than had been the case with quiet curves , and in some instances it called for considerable exercise of judgment .
To secure uniformity , it was always done by myself .
Disturbed days were those in which there was so much irregularity that smoothing appeared too arbitrary a process .
To a certain extent , no doubt , the allotment of a day to the disturbed list depended on the judge 's condition , both physical and mental , at the moment .
The attributes of a disturbed day were practically those of " character " 2 days under the international scheme 0 ( quiet ) , 1 ( moderately disturbed ) , and 2 ( highly disturbed ) ; and it is only necessary to consult the returns from similarly situated stations to recognise the importance of the personal element in the selection .
The choice , in the present case , it should be remembered , was based entirely on the D curves .
The total number of days assigned to the disturbed list in the 11 years was 209 , or au average of 19 a year .
The number varied , however , from 6 in 1890 to 39 in 1896 .
The results from the ordinary day D curves , excluding a few that were imperfect , were discussed in a paperf * 'Phil .
Trans. , ' A , vol. 202 , p. 335 .
t ' Phil. Trans. , ' A , vol. 208 , p. 205 .
FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
published in 1908 .
This paper was reprinted with the addition of an Appendix in the 'Collected Researches ' of the National Physical Laboratory , vol. 5 , 1909 .
The Appendix contains a list of the 209 disturbed days .
In 1908-9 a further grant of \#163 ; 100 from the Government Grant Committee enabled the measurement to be commenced of all the H and V curves for the period 1890 to 1900 .
A difficulty at once presented itself .
While some of the 209 days , which had been classified as disturbed from consideration of the D curves alone , were quite ordinary so far as the H curves were concerned , other days which had been treated as ordinary from the point of view of the D curves were conspicuously disturbed from the point of view of the H curves .
V is a much less disturbed element at Kew than D or H , and many of the V curves from the 209 days classified as disturbed could have been smoothed satisfactorily .
The decision reached was to regard the 209 days already selected as representing disturbed conditions for all three elements .
Diurnal inequalities were derived from these , and these only , as representative of disturbed conditions .
In their case the curves were read absolutely unsmoothed , at exact hours G.M.T. Notwithstanding the large irregularities in individual days , diurnal inequalities were obtained of a fairly regular character.* S 2 .
Coming now to ordinary days , it was decided in the case of H to set aside the 209 days already mentioned , and in addition all days when the H curves were too disturbed to smooth , and to derive inequalities from the remainder .
These were smoothed , when it seemed expedient , exactly in the same way as the D curves had been .
In the case of V a different procedure was adopted .
There was no single month in which a large majority of the curves could not be satisfactorily used without any smoothing .
This being so , it seemed best to dispense with smoothing\#151 ; which everyone admits is open to certain criticisms , while some dispute its necessity\#151 ; though that entailed omitting a considerable number of days additional to the 209 .
The discussion of D results hardly comes under the present memoir , but D enters with H into such quantities as the north and west components ( N and W ) , which of necessity are treated here .
Thus particulars of the number of days ' traces actually used for the ordinary day D inequalities concern us as well as the corresponding data for D and H. It is simpler to enumerate the days not used than those used .
This is done for individual years in Table I. , and for the 12 months of the year in Table II .
The total number of days in the 11 years , it should be remembered , was 4017 .
Natural disturbance was not the sole cause of omission of days .
A few of the days\#151 ; 10 in the case of H , 19 in the case of D and V-\#151 ; were omitted owing to imperfections in the records .
One cause of imperfection , stoppage of the clock , affected the three elements alike .
Another cause , insufficient gas supply , affected all to some extent , but while one trace might be invisible another might be measurable .
* 'Phil .
Trans. , ' A , vol. 210 , p. 271 , and 'Collected Researches , National Physical Laboratory , ' vol. 7 , p. 1 .
The H optical arrangements are the best , and it suffered least .
The erection in 1892 of a new upper story to the Observatory , with iron girders , produced irregularities on several days , which could not be satisfactorily dealt with .
There were various other discontinuities associated with movement of iron in the building , or with changes of sensitiveness in the vertical force magnetograph , which rendered the omission of certain days expedient .
But , everything considered , the number of days ' trace which could not be utilised was wonderfully small , a fact reflecting credit on the staff , especially Mr. T. W. Baker , who had charge of the magnetic instruments during the whole period concerned .
It was judged important to have a complete set of values of the absolute daily range ( maximum less minimum ) during the period .
The loss of a good many hours ' trace necessarily introduces some uncertainty into the daily range , because one at least of the extreme values might fall during the time lost ; but in many cases one can be reasonably sure that the range deduced from the part of the trace that is complete is the full range , or at least very approximately so .
In all doubtful cases recourse was had to the corresponding Falmouth curves , which were kindly lent by Mr. E. Kitto , then Superintendent of Falmouth Observatory .
In a few cases , while the ranges accepted were derived essentially from the Kew curves , a small correction was applied which was based on a comparison of the Kew and Falmouth curves .
In a few other instances the range accepted was derived from the Falmouth curve alone .
Experience showed that the agreement between Kew and Falmouth ranges was usually so close that the uncertainty thus introduced into monthly or similar mean values must be wholly negligible .
Finally , there were two or three cases in which one of the traces during a magnetic storm had gone beyond the limits of registration at the same time both at Kew and Falmouth , or the latter trace was otherwise incomplete .
In such a case there was nothing for it but to take the edge of the sheet as representing one of the extreme values .
On one occasion , February 14 , 1892 , the estimate thus made of the D range was not impossibly a very appreciable underestimate , though hardly to the extent of exercising an effect of more than a few tenths of a minute in the monthly mean .
The number of days omitted from the ordinary day diurnal inequalities through imperfections of the trace was so small that we shall not be far wrong if we disregard them when comparing different months or years in Tables I. and II .
The curves D , H or V , of the same year were dealt with at one time , while a considerable interval sometimes intervened between the consideration of two successive years .
Thus Table II .
is probably a more reliable index than Table I. to the fluctuation of disturbance in individual elements .
Since , however , the D , H and V curves were considered at widely different times , when their indications in Table I. agree the result may be accepted with some confidence .
The two quietest years were undoubtedly 1890 and 1900 , the years of lowest sunspot frequency .
The last seven months of 1900 contained no single day considered disturbed for any of the elements , and maybe accepted as the quietest period of the 11 years .
The year of largest FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
sunspot frequency , 1893 , though showing very large regular diurnal variations , was decidedly quieter than the adjacent yeai-s .
Table I.\#151 ; Number of Days not included in " Ordinary .
" Element .
1890 .
1891 .
1892 .
1893 .
1894 .
1895 .
1896 .
1897 .
1898 .
1899 .
1900 .
11 years .
D 7 26 31 17 21 21 41 16 19 2 ?
8 228 H 6 25 36 14 26 29 45 19 21 22 9 252 V 14 48 53 36 36 39 54 26 28 21 11 366 While 1892 and 1894 contained most of the outstanding magnetic storms , 1896 was remarkable for the persistence of disturbed conditions .
A point to be remembered is that it is easier to recognise the general trend of the regular diurnal variation when the range is large than when it is small .
Thus a disturbance sufficient to mask the regular diurnal variation when least\#151 ; at midwinter , in sunspot minimum\#151 ; might prove no serious obstacle to smoothing curves at midsummer near sunspot maximum .
Making all due allowance for the increased amplitude of the regular diurnal variation in summer , Table II .
shows clearly that the annual variation of disturbance has a well marked double period , with minima at midwinter and midsummer , and maxima in March ( or late February ) and October .
The maximum in Spring is the more prominent of the two .
Table II.\#151 ; Number of Days not included in " Ordinary .
" Element .
January .
February .
March .
April .
May .
June .
July .
August .
September .
October .
November .
December .
D 22 26 32 17 18 7 12 ii 19 30 20 14 H 26 26 34 19 18 8 13 18 21 29 24 16 V 40 42 45 31 28 19 23 25 30 37 26 .
20 The basis on which the curves were treated was largely determined by the fact that the D , H and Y curves had to be considered at different times .
If a careful consideration of the curves from all three elements had preceded all measurements , a common selection of ordinary days would probably have been made .
As it is , a little more disturbance enters into the ordinary day H inequality than into that for Y , and a little more into the D inequality than into that for H. This is most to be regretted perhaps in relation to the derived inequalities .
The north component ( N ) and west component ( W ) inequalities depend on both D and H , while the total force ( T ) and inclination ( I ) inequalities depend on H and Y. These inequalities , not improbably , differ slightly from what they would have been if derived from magnetographs recording N , W , T and I directly .
The differences , however , cannot he serious , because all the largest disturbances naturally figured on each of the three lists .
S 3 .
The full publication of magnetic data normally includes tables of the hourly values of three elements in absolute measure .
Besides hourly measurements of the curves , this entails for each element a knowledge of the scale value and of the base value for each ( lay .
When a temperature correction is necessary , and the temperature alters sensibly throughout the day , it entails further a knowledge of the temperature coefficient and measurements of the temperature records at each hour .
Two months ' hourly values of a single element in ordinaiy type fill a quarto page .
Thus full publication of 11 years ' data would have filled 198 quarto pages simply with the hourly values of D , H and Y. As there was no prospect of publication on this scale , and economy of effort was important , no more was done than was essential for the immediate object in view .
So long as one can assume the base value constant for the whole of each day , or can adequately allow for its fluctuation by means of a non-cyclic correction , its value is immaterial , so far as the diurnal inequality is concerned .
For simplicity , consider the case where no temperature correction is required .
If the scale value can be treated as constant for the whole of a month , as was the case at Kew with rare exceptions , all that is necessary is to take the hourly measurements in millimetres , sum the hourly columns , divide the hourly sums by the number of days in the month , allow for non-cyclic change , find the algebraic excess of each hourly mean value over the corresponding mean for the 24 hours , and convert the inequality thus formed into C.G.S. units , through multiplication by the factor representing the equivalent in force of 1 mm. When a temperature correction is required , the inequality of temperature for each month can be derived from the hourly measurements of the temperature curves .
This is converted into force from a knowledge of the temperature coefficient , and the result is applied with appropriate sign as a correction to the inequality already obtained .
There would be a great increase of labour , and no gain in accuracy , so far as the inequality is concerned , if each hourly value of the magnetic curve were corrected for temperature .
A second and conclusive reason for not correcting individual hourly values was the fact that continuous records of temperature were not taken in the magnetograph room until 1895 .
This being so , a course was followed which at least reduced labour to a minimum .
During the 11 years no change had been made in the magnetograph room , or in the programme of work done in it or adjacent parts of the building .
There was thus no reason to suspect any considerable change in the thermal phenomena in the room , and evidence pointing to the same conclusion was derivable from the 3 or 4 daily readings from mercury thermometers under the glass shades covering the H and V magnets .
It was thus decided to calculate mean diurnal inequalities of temperature , utilising the FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
thermogram measurements made on magnetic quiet days from 1895 to 1900 , and -these were taken as applying to the whole 11 years .
This attributes to each January , for instance , the same diurnal inequality of temperature , while it was no doubt larger in some Januarys than others .
This procedure no doubt introduced slight errors into the inequalities for individual months of individual years , but these would tend to disappear in results from groups of years or from the whole 11 years .
The temperature correction of the H magnetograph is about l'7y per 1 ' F. , and the range of the temperature diurnal inequality in most months was Tinder 1 ' F , so a difference of temperature lag of even 1 or 2 hours between the magnet and thermograph would have had little effect .
The fluctuations of temperature in the magnetograph room from one day to the next are sometimes much greater than the range of the regular diurnal variation , and the comparison of the readings from the thermograph and mercury thermometer on the one hand , and the corresponding fluctuation in the base values derived from individual absolute observations on the other , afforded grounds for confidence that uncorrected effects of temperature in the H inequalities must be trifling .
The Y magnet has a much larger temperature coefficient\#151 ; about 12'5y for 1 ' F.\#151 ; and the range of the regular diurnal variation is considerably less in Y than in H. Thus there is more reason to fear uncorrected temperature effect in the Y inequalities .
It is , however , mainly in the absolute daily range\#151 ; i.e. , the difference between the highest and lowest values throughout the day\#151 ; that temperature uncertainty comes in .
Undoubtedly some individual daily ranges , especially those of Y , suffered considerably from this cause .
In the case of an element like Y at Kew considerably affected by temperature , it is sometimes difficult to recognise the maximum or minimum .
If the temperature change in the day has been large , the maximum force may come at quite a different hour from the maximum ordinate .
It may be necessary to take half a dozen measurements of the force curve , with the corresponding thermogram measurements , before one can decide .
This is especially true of quiet curves\#151 ; and most Y curves are quiet\#151 ; at seasons when the regular magnetic diurnal variation is small .
No temperature corrections , of course , were possible until 1895 , and mean results based on a number of days would have been practically useless in dealing with individual days .
It was accordingly decided to attempt no temperature corrections to absolute daily ranges , but to derive them from the magnetic curves as if these required no temperature correction .
The neglect of temperature could hardly prejudice one 's view of the character of the day as quiet or disturbed , but undoubtedly in a few cases it led to a quiet day being assigned a V range more appropriate to a day of moderate disturbance .
The neglect of temperature may even have exerted a slight effect on the estimate of the mean monthly value of the absolute range , as the occurrences of maximum and minimum are much more numerous at certain hours than others , and so the effect of the regular diurnal variation of temperature would not be wholly eliminated .
Non-cyclic Change .
S 4 .
In obtaining the diurnal inequalities the non-cyelic ( n.c. ) change C\#151 ; the algebraic excess of the value at the second midnight ( hour 24 ) over that at the first midnight ( hour 0)\#151 ; has been eliminated in the usual way , by applying at hour n the correction C(l2\#151 ; ?i)/ 24 .
In the case of D the value of C for each month of the eleven years was given in a previous paper.* The value \#151 ; Cf'12 assigned there to February , 1892 , should have been \#151 ; 0''22 .
Allowing for this , the mean values for the year 1892 , for the month of February , and for the whole 11 years , become respectively \#151 ; Cf'031 , \#151 ; 0/ '052 and \#151 ; 0r,301 .
The change of force perpendicular to the magnetic meridian necessary to alter D by f varied gradually from 5'29y in Table III.\#151 ; Non-cyclic Change ( Unit O'Oly ) .
Element .
00 P 1891 .
1892 .
1893 .
1894 .
1895 .
1896 .
1897 .
1898 .
1899 .
1900 .
11 years .
D - 7 -15 - 16 -23 -52 + 4 -\#166 ; 16 - 13 -25 -21 + 9 -160 H + 2 + 68 + 200 + 69 + 90 + 84 + 118 + 80 + 63 + 61 + 42 + 79-7 V -37 -67 - 55 -55 -45 -40 - 25 -29 -30 -50 -28 -42-0 1890 to 5'36y in 1900 , the mean for the eleven years being 5'32y .
Tables III .
and IV .
give the mean yearly and monthly values of the n.c. change for the ordinary days in the three elements , those for D being expressed in terms of the equivalent force .
Ordinary days , it should be noticed , include the quiet , though , of course , not the disturbed .
Table IV.\#151 ; Non-cyclic Change ( Unit O'Oly ) .
Element .
J anuary .
February .
March .
April .
May .
June .
July .
; August .
September .
October .
November .
December .
D ' - 9 -28 - 41 - 6 + 7 - 26 - 2 0 -52 - 24 - 15 + 4 H + 41 + 97 + 202 + 75 + 49 + 6 i+ 54 + 132 + 64 + 87 + 84 + 67 Y - 115 + 11 -122 -211 - 219 -245 -126 - 2 + 98 + 214 + 141 + 73 There is obviously a prevailing tendency for the n.c. change on ordinary days to be negative in D and V and positive in H. The inference that the D and V elements were falling , and the H element rising , would be correct , though unwarranted .
The n.c. effect is influenced , of course , by the secular change but it is partly of instrumental * 'National Physical Laboratory Collected Researches , ' vol. V. ( Table IIa .
, p. 48 ) .
FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
origin , and is partly dependent on the type of day dealt with .
In the average of the Astronomer Eoyal 's quiet days , D rose 0/ -044 ( 0'23y ) , H rose 3'34y and V fell 0'84y .
Thus if all days had been quiet days , D and H would in a single year have increased respectively lG^l ( S6y ) and 1219y , while V would have fallen 307y .
The real average annual changes during the eleven years were \#151 ; 5'79 ( \#151 ; 30'8y ) in D , + 25'9y in H and \#151 ; 22'6y in V. .
It seems desirable to look into the matter a little more closely .
Take first the case of D. There were in the eleven years 209 disturbed and 19 incomplete days .
Nope of the latter , so far as could be judged , were highly disturbed .
If we regard them as ordinary , as we are fairly entitled to do , we have 3808 ordinary days .
The mean observed n.c. changes on disturbed and ordinary days were respectively + 0''327 and \#151 ; 0''0301 .
The total n.c. changes were thus + 68/ -3 on disturbed and \#151 ; 114/ -6 on ordinary days , leaving a balance of \#151 ; 46''3 .
If all measurements were exact , in the absence of instrumental change , we should expect this balance to agree with the observed secular change , but this at the observed average rate of \#151 ; 5r,79 per annum amounted to \#151 ; 63''7 .
This leaves 16/ -4 unaccounted for , suggesting an instrumental drift of l/ -5 per annum . .
In the case of H the 10 incomplete days may reasonably be regarded as ordinary , making a total of 3775 ordinary days .
The observed n.c. change on ordinary days averaging + 0*797y , the total n.c. change from the ordinary days of the eleven years conies to +3009y .
Of the remaining 242 days , 209 were included in the original list of disturbed days .
The mean n.c. change for these days was \#151 ; 13'2y , giving a total n.c. change of \#151 ; 2759y .
The remaining 33 days were included in the subsidiary list of disturbed days , which was made out when the H curves came to be treated .
The sum of the n.c. changes on these 33 days was only \#151 ; 24y .
Thus the total n.c. change for the 4017 days of the eleven years was ( + 3009 \#151 ; 2759 \#151 ; 24)y , or + 226y .
The observed secular change , +25'9 x 11 or +285y , exceeds this by only 59y , suggesting the trifling instrumental drift of \#151 ; 5'4y per annum .
In the case of V , including the incomplete amongst the ordinary days , we have 3669 of these with an average n.c. change of \#151 ; 0'420y , giving a total of \#151 ; 1541y .
The 209 disturbed days in the original list- contributed +209 x 2'7y or +564y , while the 139 disturbed days on the subsidiary list contributed +304y .
Thus for the whole 4017 days we have a balance of ( \#151 ; 1541+ 564 +304)y or \#151 ; 673y , as compared with a true secular change of \#151 ; 22'6 x lly or \#151 ; 249y .
This suggests an instrumental di'ift at the average rate of \#151 ; 39y per annum .
The chief importance perhaps of these calculations is the light they throw on the .
trustworthiness of the magnetic curves and measurements .
It is unnecessary to emphasise the fact that when instrumental creep is large it is a source of very considerable uncertainty .
The results obtained above are not put forward as exact measures of the instrumental creep , but only as showing its order of magnitude and the general fact that it was small .
Most of the quiet day curves were measured VOL. CCXVI.\#151 ; A. 2 E many years before the others , and a variety of scales were employed .
Then the curves of the 209 originally selected disturbed days were not smoothed , while those of the ordinary days were .
Thus in a good many cases , at the midnight common to a disturbed and an ordinary day two readings were taken , at widely different times , one on the unsmoothed curve , the other on a smooth pencil trace .
In individual cases these two midnight readings differed considerably , and this of course influenced the balance of the n.c. changes .
The difference between the n.c. changes in H on the 20.9 days of the original disturbed list and the supplementary list of 33 days may appear suspicious , but is easily accounted for .
During a large magnetic storm H nearly always shows a slight rise at the start .
This is usually followed by a fall , which goes on until the value has diminished below the normal , sometimes much below the normal .
There is then a recovery , which may go on at a gradually diminishing rate for some days .
The ends and beginnings of storms were represented by a larger proportion of the 33 than of the 209 days .
One of the 33 days showed an n.c. change of + 140y .
In the case of D what the absolute observations suggested was not a real instrumental drift , but occasional small discontinuities due probably to movements of iron in the building .
In the case of H there is confirmatory evidence from the base line values that the instrumental creep is in the direction simulating a fall of force , but they suggest \#151 ; 15y per annum as a more probable estimate than \#151 ; 5y as found above .
In the case of V the instrumental creep in reality seems to fluctuate in direction .
When a sensible change of sensitiveness occurred in the course of the year , the tendency to creep seemed more apparent .
On individual ordinary days the n.c. change in V is mainly a tempei'ature effect .
This may in fact be recognised in the figures given in Table IV .
The four months April to July include most of the summer rise of temperature in the magnetograph room , the principal part of the annual fall taking place in the four months September to December .
The mean daily n.c. changes during these two groups of months are by Table IV .
: From April to July \#151 ; 2'00y .
, , September to December + l'32y .
The two means will naturally include equal or approximately equal contributions from any regular source of drift which is independent of temperature , such for instance as might arise from gradual weakening of the magnet .
If we ascribe the difference between the two four months ' means obtained above solely to temperature , and take .
the known temperature coefficient , viz\lt ; \lt ; , 12'5y per 1 ' F. , then assuming the rise and fall of temperature in the two groups of months equal , we find for its amount ( 3-32/ 2 ) x ( 120/ 12-5 ) = 16-0 F. This is not far from the truth .
The annual range in reality usually exceeded 20 ' F. , but the rise usually began in February and continued throughout part of August .
FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
The fall of temperature usually continued throughout January , hut the readjustments of the magnetograph were usually made in that month leading to special uncertainties in the n.c. changes .
Diurnal Inequalities .
S 5 .
The diurnal inequality of a magnetic element is in continuous variation with the season of the year ; it also varies according to the development of sunspots , and it depends on the more or less disturbed character of the day .
There are most likely other causes of variation ; for instance , it seems unlikely that the diurnal variation at a particular station remains wholly unaffected by the secular change in the earth 's magnetism .
In deciding on the amount of detail advisable in the presentation of the facts , several conflicting considerations have to be allowed for .
There is a great deal which in the present state of our knowledge must be regarded as accidental in the magnetic changes on any individual day .
If we derive inequalities from the combination of a very limited number of days , a good deal of this " accidental " element will remain uneliminated .
If , on the other hand , we combine a large number of days from the same year , the inequality is inevitably a blend of more or less conflicting characteristics .
The extent to which this is the case differs at different seasons of the year .
There is , for instance , much greater variation in the type and amplitude of the diurnal inequality in the five months November to March than in the five months April to August .
If we have at our disposal the data from a large number of years , we can get smooth diurnal inequalities for individual seasons of the year a good deal shorter than a calendar month .
This is what we should naturally do if our object were to examine in very minute detail the mode of variation of the diurnal inequality throughout the year .
It is , however , open to the objection that it would produce a mass of detail which few readers if any could digest .
The object in view moreover might be to some extent defeated by the influence of the sunspot relationship and secular change effects .
Considerations of space must also be borne in mind .
Diurnal inequalities for each of the 132 months of the eleven years\#151 ; let alone shorter periods of the year\#151 ; for H , V , N , W , T and I would have entailed printing an immense mass of figures .
As a matter of fact , diurnal inequalities were calculated for TI and V from each of the 132 months , but not for the other elements .
The H and V ranges from these 132 inequalities are given later , but it was decided to publish for each of these elements only four tables of diurnal inequalities .
Thus , in the case of H , Table V. gives the 12 diurnal inequalities obtained by combining all the months of the same name in the 11 years , three diurnal inequalities for the seasons winter ( November to February ) , equinox ( March , April , September and October ) and summer ( May to August ) , and finally the mean diurnal inequality for the whole year .
Table VI .
differs from Table V. only in that it is confined to the four years 1892 to 1895 representing large sunspot frequency ; for brevity , it is described as referring to sunspot maximum .
Table VII .
is similarly confined to 1890 , 1899 , and 1900 , and described as representing sunspot minimum .
The mean values of Wolfer 's sunspot frequency for the years included in these three tables were respectively 41'7 , 75'0 and 7'2 .
Finally Table VIII .
gives for each of the eleven years the mean diurnal inequality for the whole year .
Tables IX .
to XII .
for V correspond exactly with the four tables for H. The n.c. element has been eliminated from all these tables in the way already described .
All the inequalities refer to G.M.T. Kew local time is only l|- minutes after Greenwich , so the employment of local time would have made little difference .
S 6 .
The other inequalities were calculated from the H , V and D inequalities , taking the latter as given in my previous paper .
We have T2 = H2 + V2 , tan I = V/ H , N = HcosD , W = HsinD .
( 1 ) .
Thus if AT , AH , \amp ; c. , represent corresponding small departures from the mean values we have AT = cos I AH + sin IAV , AI = isin 21 ( AV/ V-AH/ H ) , AN = cos D AH \#151 ; H sin D AD , AWE = sin D AH + H cos D AD ( 2 ) .
The coefficients of AH , AV , \amp ; c. , on the right-hand sides of these equations varied slightly throughout the eleven years , in consequence of the secular change .
They were treated as constant throughout individual years , and mean values were taken for inequalities based on the eleven years .
In the latter case the formulae actually used were AT = 0-384 AH + 0'923 AV , AI = 0'02775 AV\#151 ; 0"0668 AH , AN = 0-955 AH-U58 AD , AWE = 0-298 AH + 5'08 AD ( 3 ) .
where the unit is 1 ' in the case of AI and AD , and lie in the case of the force elements .
Table XIII .
gives diurnal inequalities for T. The results for the 12 months are from the 11 years combined .
But in addition to a diurnal inequality for the year from the 11 years combined , there are corresponding inequalities for the sunspot maximum and minimum groups of years .
For I two inequality tables are given .
Table XIV .
contains inequalities for the 12 months and the whole year from the 11 years combined , being derived from Tables V. and IX .
It also contains diurnal inequalities for the year from the sunspot maximum and minimum groups of years .
Table XV .
gives diurnal inequalities for the whole year , from the separate years , based on Tables VIII .
and XII .
Tables XVI .
and XVII .
give monthly and seasonal diurnal inequalities for N and W FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
from the whole 11 years , and diurnal inequalities for the year from the 11 years and the groups of years of sunspot maximum and minimum .
To save decimals O'ly is employed as the unit in the tables relating to elements of force .
The extreme hourly values , i.e. , the algebraic maximum and minimum , are in heavy type .
In addition to the hourly values , the tables give the range or algebraic difference of the extreme hourly values , and the quantity described as A.D. ( or average departure from the daily mean ) .
The latter quantity represents the result obtained by dividing by 24 the numerical sum of the differences of the 24 hourly values from their arithmetic mean .
If readings were taken at every minute , instead of every hour of the day , larger values would in most cases be obtained for the range of the inequality , because it must be exceptional for the extreme values to fall at exact hours G.M.T. The underestimate , however , thus arising is usually very small , as is easily recognised from the shape of the curves representing the inequalities .
The value of the A.D. is naturally less affected by the accident of time , and in most cases it probably gives a better idea than the range of the activity of the forces to which the diurnal inequality is due .
This is more especially the case when the inequality shows a double daily variation with two maxima and two minima .
S 7 .
The inequality data in Table V. are shown graphically in fig. 1 .
As in other similar cases , the general features are most readily recognised in the curve , while for details recourse is desirable to the numerical data .
Fig. 2 shows the H inequalities for the three seasons and the year , contrasting the data for the whole eleven years in Table V. with those for the sunspot maximum and minimum groups of years in Tables VI .
and VII .
The composite character of the diurnal inequality in H derived from the whole year is most clearly seen by comparing some of the seasonal data in the morning hours in Table V. At 6h .
, for instance , while the winter value has its maximum for the day , the summer value falls below the daily mean .
The transition from plus to minus in winter does not occur until nearly 9h .
, while it occurs in equinox shortly after 7 a.m. , and in summer shortly after 5 a.m. The minimum or principal minimum for the day occurs at lOh .
from May to September , and at 11h .
in the remaining seven months .
It is the most constant and dominant feature in the inequality .
In the four summer months the afternoon maximum\#151 ; then the only maximum\#151 ; is almost equally prominent , but in the other months the afternoon maximum resembles a plateau rather than a peak , and in the winter months the principal maximum occurs in the forenoon .
The forenoon maximum exists also as a secondary maximum in March and October , and even at midsummer the appearance of the curves suggests some influence delaying the plunge to the minimum at lOh .
In most months the rise to the afternoon maximum seems to lag somewhat near 4 p.m. , and a distinct secondary minimum is then recognisable in the case of January .
Table V.\#151 ; Diurnal Inequality in Horizontal lb. 2h .
3h .
4h .
5h .
6h .
7h .
8h .
9h .
loll .
lib. 12b .
January . .
+ 3 4 7 4 19 4 35 + 57 4 TO + 67 + 39 - 17 - 63 - m - 70 February . .
4 17 + 14 4 16 4 30 4 50 4 67 4 69 4 41 - 21 - 78 -112 - 93 March . . .
+ 46 + 43 4 45 4 45 + 58 4 63 4 44 - 16 -109 -181 -204 -153 April . . .
4 74 4 56 4 50 4 47 4 46 4 40 - 2 - 78 -178 -262 -271 -202 May ... .
4 60 4 46 4 36 4 26 ~ 4 8 - 31 - 97 -164 -213 -235 -214 -153 June ... 4 52 4 38 4 33 4 28 + 6 - 47 - 115 -181 -235 -255 -224 -162 July . . . .
+ 56 4 42 + 32 + 25 4 2 - 44 - 107 -175 -235 -262 -238 -167 August . . .
4 81 + 62 + 57 4 41 4 15 - 23 - 98 -183 -247 -266 -225 -137 September . .
+ 78 4 67 4 6L 4 61 + 55 + 26 - 32 -117 -202 -247 -227 -135 October . .
4 62 4 60 4 64 + 74 4 86 4 80 + 49 - 26 -130 -207 -213 -165 November . .
4 18 4 22 4 34 4 48 4 65 4 T5 + 65 + 27 - 48 -105 -122 - 94 December . .
- 2 4 3 + 17 + 34 4 56 4 72 + 67 + 48 4 6 - 41 - 64 - 63 Winter . .
# .
4 9 4 12 4 22 + 37 4 57 4 71 + 67 4 39 - 20 - 72 - 97 - 80 Equinox . .
4 65 + 57 + 55 + 57 4 61 + 52 + 15 - 59 -155 -224 -229 -164 Summer . .
+ 62 4 47 4 39 4 30 + 8 - 36 \#151 ; 104 -176 -233 -255 -225 -155 Year . . .
4 45 4 38 4 39 + 41 4 42 + 29 \#151 ; 8 - 65 -136 -184 -184 -133 Table YI.\#151 ; Diurnal Inequality in Horizontal Force , lh .
2b .
3h .
4b .
5h .
6h .
7b .
8b .
9b .
10h .
11h .
12b .
January . .
4 21 4 24 + 36 4 52 4 71 4 82 + 74 4 38 - 30 - 93 -125 -Ill February . .
+ 24 + 23 + 25 4 39 4 66 4 86 4 81 4 39 - 38 -107 -149 -128 March . . .
+ 55 + 55 4 62 4 64 4 74 4 77 4 43 - 27 -137 -231 -265 -197 April . . .
4 97 4 72 4 64 4 58 4 52 4 45 - 15 -104 -224 -332 -335 -250 May ... .
4 70 + 50 4 45 4 34 4 13 - 27 - 104 -188 -254 \#151 ; 285 -269 -194 J un . . .
4 61 + 46 4 41 4 36 4 8 - 59 - 149 -241 -311 -324 -275 -194 July . . . .
+ 69 + 55 4 38 4 29 - 6 - 61 - 132 -212 -288 -328 -307 -219 August . . .
+ 99 + 77 4 72 4 56 4 16 - 21 - 115 -215 -292 -318 -282 -188 September . .
4 95 4- 84 4 73 4 72 4 64 4 34 - 33 -133 -233 -269 -253 -163 October . .
4 82 + 84 4 89 4 99 111 4 100 4 59 - 34 -159 -261 -272 \#151 ; 212 November . .
4 35 + 46 4 58 4 64 4 81 4 93 4 81 4 29 - 65 -131 -163 -137 December . .
4 9 + 10 4 24 4 44 4 67 4 84 4 74 4 53 0 - 58 - 94 - 95 Winter . . .
+ 22 + 26 4 36 4 50 4 71 4 86 4 77 4 40 - 33 - 97 -133 -118 Equinox . .
4 82 + 74 4 72 4 73 4 75 4 64 4 13 - 75 -188 -273 -281 -205 Summer . .
4 75 + 57 4 49 4 39 4 8 - 42 - 125 -214 -286 -314 -283 -199 Year . . .
+ 60 + 52 4 52 4 54 4 51 4 36 \#151 ; 11 - 83 -169 -228 -232 -174 FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
Force , from 11 Years .
( Unit O'ly .
) 13h .
14h .
15h .
16h .
I7h .
181i .
191i .
20h .
211i .
22h .
23h .
24b .
Range .
A.I ) .
- 29 5 \#151 ; 7 _ 16 \#151 ; 12 0 4 5 + 4 0 + 1 + 1 + 1 160 25 8 - 57 - 27 - 16 - 15 - 10 4 1 4 10 + 21 + 25 + 24 + 25 + 21 181 35 *8 - 87 - 32 + 6 + 2L 4 23 4 37 4- 53 + 60 + 62 + 63 + 60 + 53 267 65 2 -125 - 47 + 10 + 50 + 83 4-105 4113 4 110 + 103 + 98 + 93 + 89 383 97 1 - 93 - 32 + 25 + 74 4 123 4 154 + 161 + 143 + 120 + 100 + 86 + 72 399 102 9 - 95 - 21 + 46 + 90 4-133 4 165 + 183 + 168 + 136 + 107 + 83 + 67 438 111-3 - 91 \#151 ; 14 + 53 + 95 4-130 4 156 + 169 4 162 + 137 + 111 + 90 + 71 431 111 0 - 58 + 1 + 40 + 59 4 80 4 110 + 136 + 136 + 127 + 111 + 95 + 87 402 103 1 - 58 - 14 4 5 + 18 4 35 4 64 + 89 + 97 + 95 + 96 + 98 + 86 345 86 0 -103 - 51 - 19 - 6 4 16 -4 40 + 53 + 60 + 65 + 71 + 72 + 68 299 76 -7 - 62 - 40 - 26 - 8 + 12 -4 18 + 21 + 22 + 20 + 20 + 18 + 19 197 42 0 - 43 30 \#151 ; 23 \#151 ; 15 \#151 ; 5 - 2 - 1 \#151 ; 5 - 4 \#151 ; 3 \#151 ; 1 \#151 ; 2 136 25 -3 - 48 _ 26 \#151 ; 18 _ 14 \#151 ; 4 4 4 + 9 + 10 + 10 + 11 + 11 + 10 168 31 -6 - 93 - 36 0 + 21 4 39 4 61 + 77 + 82 + 81 + 83 + 81 + 74 311 80 -0 - 84 - 17 + 41 + 80 + 117 4 146 + 163 + 152 + 130 + 107 + 89 + 74 418 107*1 - 75 \#151 ; 26 + 8 4 29 .4 51 4 71 + 83 + 81 + 74 + 67 + 60 + 53 267 67 *6 from Sunspot Maximum Years .
( Unit O'ly .
) 13h .
14h .
15h .
16h .
I7h .
18b .
19h .
20b .
211 ) .
22h .
23h .
24h .
Range .
A.D. - 66 \#151 ; 28 \#151 ; 22 - 23 \#151 ; 9 + 8 + 15 + 16 + 18 + 19 + 18 + 15 207 42 *2 - 87 52 - 25 - 10 + 5 + 18 + 22 + 30 + 35 + 34 + 39 + 30 235 49 *7 -115 - 38 + 10 + 30 + 35 + 56 + 71 + 77 + 80 + 78 + 77 + 66 345 84*2 -154 - 64 + 6 + 64 + 102 + 132 + 149 + 147 + 133 + 128 + 117 + 112 484 123 *2 -124 - 49 + 24 + 88 + 149 + 188 + 201 + 176 + 151 + 125 + 100 + 82 486 124 -6 -109 15 + 63 + 126 + 175 + 210 + 226 + 207 + 168 + 129 + 97 + 79 550 139 -7 -124 27 + 64 + 125 + 179 + 209 + 219 + 202 + 171 + 137 + 115 + 92 547 142 -0 - 98 - 8 + 48 + 78 + 103 + 137 + 167 + 168 + 158 + 137 + 117 + 105 486 128 *1 - 81 - 32 5 + 17 + 37 + 74 + 102 + 112 + 109 + 113 + 114 + 100 383 100 -i -135 - 71 - 25 - 1 + 23 + 52 + 63 + 75 + 81 + 84 + 84 + 85 383 97 *5 -106 - 67 40 - 16 + 8 + 19 + 29 + 33 + 42 + 40 + 34 + 33 256 60 *4 - 74 \#151 ; 52 \#151 ; 35 - 15 - 2 + 2 + 4 + 8 + 8 + 13 + 13 + 12 179 35*4 - 83 \#151 ; 50 \#151 ; 30 - 16 0 + 12 + 17 + 22 + 26 + 27 + 26 + 22 219 46 7 -121 51 4 + 28 + 49 + 78 + 96 + 103 + 101 + 101 + 98 + 91 384 99-8 -114 - 25 + 51 +104 ' + 151 + 186 + 203 + 188 + 162 + 132 + 107 + 90 517 133 o -106 \#151 ; 42 + 6 + 39 + 67 + 92 + 106 + 104 + 96 + 86 + 77 + 68 338 87 -1 O Z Z 6 8 / O M\gt ; CIl 14- 16 18 SO ZZ Z4 . .
* \#187 ; - I II I i . .
Jan Feb March June Fig. 1 .
Horizontal force .
11 years .
FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
0 Z 4- 6 8 1C Noon 16 is so Equinox YOL .
CCXVI.\#151 ; A. Fig , 2 .
Horizontal force .
2 F Table VII.\#151 ; Diurnal Inequality in Horizontal Force , lb. 2b .
3h .
4h .
5b .
6h .
7b .
8b .
9b .
10h .
lib. 12h .
January . .
\#151 ; 16 \#151 ; 17 \#151 ; 5 + 9 + 32 + 47 + 49 + 31 - 12 - 30 - 47 - 17 February . .
+ 16 + 7 + 10 + 20 + 34 + 48 + 53 + 36 - 20 - 61 - 88 - 69 March . . .
+ 31 + 27 + 29 + 27 + 41 + 48 + 41 - 10 - 89 -143 -157 -112 April . . .
+ 55 + 37 + 29 + 27 + 28 + 26 + 5 - 47 -129 -195 -198 -145 May ... .
+ 51 + 42 + 33 + 23 + 7 - 30 - 84 - 137 -175 -187 -160 -108 June . . .
+ 45 + 31 + 24 + 27 + 15 - 22 - 76 - 132 -182 -208 -185 -134 July . . . .
+ 48 + 32 + 22 + 18 + 5 - 32 - 89 - 146 -194 -204 -173 -104 August . . .
63 + 53 + 44 4- 30 + 11 - 27 - 87 - 154 -201 -213 -167 - 86 September .
+ 55 .
+ 46 + 43 + 49 + 50 + 24 - 25 - 94 -168 -216 -195 -105 October . .
+ 42 + 43 + 47 + 55 + 68 + 63 + 39 - 21 - 98 -154 -159 -127 November . .
+ 5 4 + 18 + 35 + 53 + 58 + 44 + 17 - 41 - 84 - 87 - 65 December . .
\#151 ; 7 \#151 ; 1 + 12 + 29 + 50 + 63 + 62 + 47 + 7 - 33 - 45 - 39 Winter .
, .
\#151 ; 1 \#151 ; 2 + 9 + 23 + 42 + 54 + 52 + 33 - 17 - 52 - 67 - 48 Equinox . .
+ 46 + 38 + 37 + 40 + 47 + 40 + 15 - 43 -121 -177 -177 -122 Summer . .
+ 52 + 39 + 31 + 24 + 9 - 28 - 84 \#151 ; 142 -188 -203 -171 -108 Year . . .
+ 32 + 25 + 25 + 29 + 33 + 22 \#151 ; 6 \#151 ; 51 -108 -144 -138 - 93 Table VIII.\#151 ; Diurnal Inequality in Horizontal Year .
Hi .
2h .
3h .
4h .
5b .
6b .
7h .
8b .
9h .
10h .
11h .
12h .
1890 + 34 + 25 + 24 + 29 + 33 + 19 __ 10 - 52 -109 -142 -133 - 84 1891 + 42 + 37 -f 38 + 43 + 43 + 25 - 11 - 67 -135 -182 -182 -134 1892 + 61 + 51 + 49 + 49 + 39 + 26 - 18 - 89 -174 -232 -231 -166 1893 + 64 + 56 + 55 + 59 + 62 + 41 - 9 - 86 -178 -244 -251 -187 1894 * + 60 + 51 + 55 + 55 + 54 + 40 - 8 - 82 -169 -230 -234 -186 1895 + 55 + 51 + 49 + 52 + 51 + 37 - 10 - 75 -156 -206 -213 -157 1896 + 44 + 36 + 36 + 36 + 41 + 29 - 3 - 56 -129 -180 -184 -135 1897 + 41 + 33 + 33 + 37 + 39 + 32 + 3 - 50 -110 -160 -159 -114 1898 + 38 + 32 + 31 + 34 + 35 + 21 - 10 - 61 -117 -132 -149 -105 1899 + 36 + 28 + 27 + 30 + 35 + 26 - 5 \#151 ; 55 -117 -154 -147 -102 1900 + 27 + 23 + 25 + 29 + 31 + 22 \#151 ; 3 - 46 -100 -137 -136 - 92 FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
from Sunspot Minimum Years .
( Unit O'ly .
) 13h .
14k .
15h .
16b .
17b .
18h .
19h .
20h .
21h .
22k .
23h .
24h .
Range .
A.D. + 20 + 32 + 21 + 1 \#151 ; 17 \#151 ; 10 \#151 ; 9 \#151 ; 9 \#151 ; 14 \#151 ; 14 \#151 ; 12 \#151 ; 11 96 20 -1 \#151 ; 30 2 + 1 - 5 - 10 - 7 + 2 + 11 + 15 + 10 + 15 + 17 141 24 *5 \#151 ; 55 \#151 ; 13 + 17 + 23 + 17 + 24 + 34 *+ 43 + 44 + 47 + 44 + 41 205 48 *2 \#151 ; 92 - 27 + 22 + 41 + 60 + 70 + 15 + 15 + 15 + 71 + 68 + 69 273 69 4 \#151 ; 58 11 + 26 + 55 + 89 +114 + 123 + 105 + 92 + 76 + 66 + 54 310 79 *4 \#151 ; 72 \#151 ; 13 + 31 + 60 + 89 + 119 + 140 + 128 + 106 + 86 + 68 + 55 348 85 *3 \#151 ; 48 + 5 + 42 + 61 + 78 + 104 + 123 + 123 + 103 + 90 + 73 + 61 327 82 *4 \#151 ; 22 + 15 + 34 + .
39 + 46 + 76 + 108 + 105 + 101 + 91 + 77 + 70 321 80 -0 - 35 - 1 + 13 + 19 + 31 + 55 + 79 + 80 + 74 + 74 + 80 .
+ 68 296 70 -0 \#151 ; 79 - 38 14 - 2 + 16 + 35 + 42 + 42 + 47 + 53 + 50 + 46 227 57*5 - 32 16 8 + 6 + 17 + 21 + 19 + 11 + 6 + 6 + 7 + 8 145 27 *8 \#151 ; 14 \#151 ; 11 \#151 ; 13 \#151 ; 11 \#151 ; 4 3 \#151 ; 13 \#151 ; 23 \#151 ; 19 \#151 ; 15 \#151 ; 9 \#151 ; 9 108 22 *5 \#151 ; 14 + 1 0 \#151 ; 2 _ 3 0 0 \#151 ; 3 \#151 ; 3 \#151 ; 3 0 + 1 121 n -9 - 65 - 20 + 10 + 20 + 31 + 46 + 58 + 60 + 60 + 61 + 60 + 56 238 60 4 - 50 | ~ 1 + 33 + 54 + 75 + 103 123 + 115 + 100 + 86 + 71 + 60 326 81 '3 \#151 ; 43 \#151 ; 7 + 14 + 24 + 34 + 50 + GO + 58 + 53 + 48 + 41 + 39 204 49-2 Force , from Individual Years .
( Unit O'ly .
) 13li .
14b .
lob .
16k .
17k .
18k .
19k .
20k .
21h .
22h .
23k .
24k .
Range .
A.D. - 34 0 + 17 + 29 + 34 + 48 + 56 + 51 + 47 + 41 + 39 + 38 198 47 0 - 73 - 20 + 18 + 35 + 58 1- 76 + 80 + 76 + 66 + 60 + 57 + 50 262 67 *0 -100 - 36 + 10 + 43 + 72 + 95 + 101 + 105 + 99 + 89 + 81 + 71 339 87 *2 -110 - 38 + 14 + 47 + 71 + 92 + 105 + 104 + 95 + 87 + 79 + 73 356 92 *0 -116 - 53 - 4 + 35 + 70 + 10l + 114 + 112 + 102 + 89 + 78 + 66 348 90 -2 - 97 - 41 + 2 + 30 + 55 + 81 + 91 + 96 + 88 + 80 + 71 + 61 310 79 *6 - 77 - 29 + 6 + 27 + 48 + 69 + 82 + 84 + 75 + 68 + 59 + 53 268 66 -1 - 67 - 32 \#151 ; 6 + 12 + 40 + 62 + 16 + 74 + 63 + 56 + 53 + 43 236 58 *1 - 57 - 18 + 3 + 18 + 39 + 54 + 71 + 12 + 66 + 58 + 52 + 45 224 55 8 - 55 - 16 + 8 + 19 + 36 + 54 + GG + 65 + 61 + 59 + 55 + 47 220 54 3 - 41 - 4 + 17 + 24 + 33 + 48 + 59 + 57 + 50 + 44 + 38 + 32 196 46*6 Table IX.\#151 ; Diurnal Inequality in Vertical lh .
2h .
3b .
4b .
5b .
6h .
7h .
8h .
9h .
lOh .
lib. 12b .
January .
_ 6 10 _ 16 _ 18 \#151 ; 19 \#151 ; 20 - 23 - 25 - 30 - 36 - 33 - 39 February .
- 8 - 20 - 24 - 19 - 21 - 23 - 23 - 23 - 31 - 61 - TO - 61 March . .
- 6 - 15 - 21 - 20 - 17 - 15 - 6 - 10 - 44 - 90 -126 -127 April . .
+ 4 - 8 - 14 - 15 - 8 + 2 + 10 - 4 - 51 -106 -153 -167 May . . .
+ 7 0 - 2 + 2 + 7 + 6 - 10 - 35 - 85 -149 -194 -190 J un . .
\#151 ; 5 - 19 - 20 - 14 - 8 - 16 - 25 - 51 - 84 -129 -162 -154 July . . .
- 3 - 25 - 33 - 30 - 21 - 34 - 41 - 55 - 85 -115 -151 -155 August . .
- 2 - 10 - 10 - 8 + 4 + 9 + 6 - 13 - 59 -106 -143 -147 September .
- 2 - 9 - 16 - 18 - 21 - 15 - 11 - 28 - 66 -106 \#151 ; 127 -122 October .
- 8 - 14 - 23 - 24 - 26 - 25 - 15 - 10 - 31 - 72 - 98 - 89 November .
- 15 - 20 \#151 ; 23 - 23 - 22 - 27 - 31 - 25 - 31 - 54 - 54 - 37 December .
\#151 ; 3 - 12 - 15 \#151 ; 17 - 13 - 12 - 13 - 16 - 23 - 36 - 34 - 35 Winter . .
_ 8 16 _ 20 _ 19 \#151 ; 19 \#151 ; 20 - 23 - 22 - 29 - 47 - 48 - 43 Equinox .
- 3 - 12 - 18 - 19 - 18 - 13 - 6 - 13 - 48 - 94 -126 -126 Summer .
- 1 - 13 - 16 \#151 ; 13 - 5 - 9 - 17 - 38 - 78 -125 -163 -162 Y ear . . .
\#151 ; 4 - 14 \#151 ; 18 \#151 ; 17 \#151 ; 14 \#151 ; 14 - 15 - 25 - 52 - 88 -112 -110 Table X.\#151 ; Diurnal Inequality in Vertical Force , lb. 2h .
3h .
4b .
5b .
6b .
7h .
8h .
9k .
10k .
llli .
12k .
January . .
- 4 - 10 - 20 - 24 25 \#151 ; 24 _ 27 28 - 34 - 45 - 41 - 50 February . .
- 6 - 21 - 24 - 20 - 26 - 31 - 27 - 23 - 37 - 78 - 88 - 74 March . . .
-17 - 22 - 34 - 30 - 22 - 18 0 0 - 38 - 90 -130 -133 April . . .
- 5 - 18 - 22 - 21 - 8 + 8 + 20 + 7 - 46 -105 -160 -178 May ... .
0 - 5 - 7 + 3 + 13 + 13 - 5 - 35 - 91 -168 -223 -215 June . . .
- 16 - 28 - 28 - 18 \#151 ; 6 - 10 \#151 ; 20 \#151 ; 53 - 89 -146 -185 -180 July . . . .
- 9 - 39 - 49 - 47 \#151 ; 36 - 45 - 56 - 73 -102 -135 -171 -178 August . . .
- 5 - 13 - 15 - 13 + 2 + 10 + 8 - 10 - 59 -115 -153 -156 September . .
- 5 - 11 - 18 - 23 - 26 - 19 - 13 - 31 - 73 -116 -138 -130 October . .
- 12 - 22 - 34 - 36 - 39 - 38 - 24 - 16 - 31 - 83 -112 -105 November . .
- 21 - 28 - 30 - 31 \#151 ; 29 \#151 ; 33 - 36 - 24 - 30 - 59 - 62 - 45 December . .
- 12 - 25 - 26 - 26 - 22 \#151 ; 19 \#151 ; 19 \#151 ; 19 - 26 - 43 - 46 - 42 Winter . . .
- 11 - 21 - 25 - 25 26 27 27 24 - 32 - 56 - 59 - 53 Equinox .
* .
- 10 - 18 - 27 - 28 - 24 \#151 ; 17 \#151 ; 4 - 10 - 48 - 99 -135 -136 Summer . .
- 8 - 21 - 25 - 19 \#151 ; 7 \#151 ; 8 \#151 ; 18 - 43 - 85 -141 -183 -182 Year . . . .
- 9 - 20 -26 - 24 - 19 - 17 - 17 \#151 ; 25 - 55 - 99 -126 -124 FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
Force , from 11 Years .
( Unit O'ly .
) 131i .
14b .
15h .
16h .
l7h .
18h .
19h .
20h .
21h .
22k .
23b .
24h .
Range .
A.D. - 23 + 13 + 33 + 34 + 40 + 39 + 38 + 38 + 33 + 18 + 10 + 2 79 24 *8 - 44 \#151 ; 4 + 38 + 60 + 68 + 66 + 57 + 53 + 41 + 27 + 16 + 6 138 36 *0 - 93 \#151 ; 30 + 38 + 84 + 103 + 98 + 89 + 78 + 63 + 44 + 19 + 2 230 51 6 -132 \#151 ; 51 + 16 + 59 + 96 + 116 + 310 + 97 + 77 + 58 + 40 + 23 283 59 -0 -138 - 61 + 9 + 67 + 115 + 142 + 139 + 125 + 98 + 70 + 50 + 28 336 72 -0 -109 \#151 ; 43 + 17 + 75 + 120 + 141 + 142 + 127 + 98 + 64 + 37 + 19 304 70 *0 -120 - 50 + 24 + 81 + 128 + 154 + 149 + 134 + 106 + 75 + 45 + 22 309 76 *5 -107 \#151 ; 38 + 31 + 74 + 101 + 106 + 89 + 80 + 63 + 45 + 25 + 9 253 53 5 - 78 \#151 ; 16 + 39 + 81 + 94 + 94 + 93 + 82 + 67 + 50 + 30 + 8 221 53 0 - 57 \#151 ; 11 + 38 + 69 + T2 + 72 + 67 + 65 + 57 + 38 + 20 + 6 170 42 *0 - 15 + 29 + 53 + 61 + 58 + 52 + 46 + 38 + 33 + 15 + 2 - 9 115 32 2 - 22 + 5 + 21 + 32 + 35 + 34 + 31 + 29 * + 28 + 21 + 13 + 5 71 21 0 j - 26 + 11 + 36 + 47 + 50 + 48 + 43 + 40 + 34 + 20 + 10 + 1 98 28-3 - 90 - 27 + 33 + 73 + 91 + 95 + 90 + 81 + 66 + 47 + 27 + 10 221 51 1 -118 - 48 + 20 + 74 + 116 + 136 + 130 + 117 + 91 + 64 + 39 + 19 299 67 '2 - 78 \#151 ; 21 + 30 + 65 + 86 + 93 + 87 + 79 + 64 + 44 + 26 + 10 205 48 6 from Sunspot Maximum Years .
( Unit O'ly .
) 13h .
14h .
15h .
16h .
17h .
18h .
19h .
20h .
21h .
22b .
23h .
24h .
Range .
A.D. - 35 + 9 + 34 + 43 + 49 + 49 + 51 + 49 + 44 + 23 + 12 + 4 101 30 '6 - 52 - 3 + 45 + 69 + 82 + 77 + 71 + 64 + 47 + 29 + 19 + 7 170 42 *5 - 98 - 30 + 44 + 97 + 121 + 117 +100 + 85 + 63 + 38 + 7 - 10 254 56 *0 -145 - 57 + 13 + 67 +107 + 126 + 120 + 106 + 83 + 57 + 37 + 14 304 63 8 -159 - 78 + 6 + 72 + 135 + 113 + 164 + 146 + 110 + 76 + 51 + 24 396 82 *2 -129 - 51 + 24 + 89 + 146 + 169 + 164 + 142 + 109 + 68 + 35 + 13 354 79 9 -141 - 53 + 34 + 104 + 162 + 193 + 184 + 164 + 127 + 89 + 52 + 25 371 94 *5 \#151 ; 117 - 36 + 34 + 86 + 114 + 114 + 95 + 84 + 64 + 46 + 24 + 11 270 57 *7 - 82 - 16 + 48 + 91 +107 + 105 + 102 + 89 + 72 + 52 + 30 + 5 245 58 *4 - 68 - 13 + 51 + 84 + 91 + 95 + 86 + 81 + 70 + 43 + 23 + 6 209 53 *0 - 24 + 31 + 63 + 81 + 75 + 65 + 55 + 46 + 35 + 11 0 - 10 143 38 *5 - 27 + 11 + 33 + 48 + 50 + 45 + 44 + 44 + 37 + 25 + 13 2 96 29 *3 - 34 + 12 + 44 + 60 + 64 + 59 + 55 + 51 + 41 + 22 + 11 + 1 123 35 0 - 98 - 29 + 39 + 85 + 108 + 111 + 102 + 90 + 72 + 48 + 24 + 4 247 56 9 -137 - 55 + 25 + 88 + 139 + 162 + 152 + 134 + 103 + 70 + 41 + 18 345 77 *7 - 90 - 24 + 36 + 78 + 104 + 111 + 103 + 92 + 72 + 46 + 25 + 8 237 56 *2 10 Noon / 4 16 18 20 Z4J)ce , Fig. 3 .
Vertical force .
FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
o ft f 6 8 to Mon w t6 / \amp ; zz / / Yea 'unspot Yccu* Midnt Noon / s Mixing Fig. 4 .
Vertical force .
Table XI.\#151 ; Diurnal Inequality in Vertical Force , lh .
2h .
3h .
4h .
5h .
6h .
7h .
8h .
9h .
lOh .
11h .
12h .
January .
9 \#151 ; 11 \#151 ; 14 \#151 ; 13 \#151 ; 11 \#151 ; 13 \#151 ; 17 \#151 ; 21 \#151 ; 26 - 30 - 25 - 28 February .
- 8 - 16 - 21 - 18 - 16 - 16 - 17 - 20 - 25 - 43 - 49 - 46 March . .
+ 8 - 2 - 6 - 5 6 - 8 - 2 - 7 - 35 - 75 -108 -113 April . .
+ 17 + 7 + 3 + 3 2 + 5 .
+ 7 - 9 - 50 - 99 -140 -154 May . . .
+ 18 + 10 + 2 + 8 + 9 + 3 - 12 - 34 - 80 -127 -162 -165 June . .
+ 2 \#151 ; 10 - 9 - 4 - 4 - 14 - 20 - 39 - 70 -111 -140 -131 July . . .
- 7 - 24 - 28 - 24 - 17 - 31 - 32 - 42 - 68 - 92 \#151 ; 127 -122 August ) . .
+ 3 - 1 - 2 + 3 + 15 + 15 + 12 - 6 - 48 - 86 -119 -126 September .
0 - 5 - 9 - 8 - 11 - 8 - 3 - 19 - 52 - 86 -108 -105 October . .
- 1 - 7 - 14 - 12 - 12 - 11 \#151 ; 5 - 2 - 17 - 50 - 74 - 74 November .
- 8 - 9 - 11 - 15 ~ 13 - 19 - 20 - 21 \#151 ; 26 - 43 - 42 - 28 December .
+ 2 - 3 \#151 ; 6 4 \#151 ; 4 - 5 \#151 ; 9 \#151 ; 16 - 29 - 25 - 27 Winter . .
6 10 \#151 ; 13 \#151 ; 13 _ 11 13 _ 15 _ 18 23 - 36 - 35 - 32 Equinox .
+ 6 - 2 - 7 - 6 7 - 5 - 1 - 9 - 39 - 78 -107 -111 Summer .
+ 4 - 6 - 9 - 4 + 1 - 7 - 13 - 30 67 -104 -137 -136 Year . . .
+ 1 \#151 ; 6 \#151 ; 9 \#151 ; 8 \#151 ; 6 \#151 ; 8 \#151 ; 10 \#151 ; 19 \#151 ; 43 - 73 - 93 - 93 Table XII.\#151 ; Diurnal Inequality in Vertical Year .
lh .
2h .
3h .
4h .
5h .
6h .
7h .
8h .
9h .
lOh .
11h .
12h .
1890 0 _ , .
8 \#151 ; 12 - 8 \#151 ; 5 8 - 8 - 18 - 40 -.68 - 86 - 87 1891 - 9 - 13 - 19 - 19 - 14 \#151 ; 13 - 14 - 23 - 51 - 88 \#151 ; 112 -110 1892 - 16 - 25 - 32 - 28 - 21 - 21 - 22 - 32 - 57 - 95 -123 -122 1893 - 4 - 15 - 17 - 13 - 7 - 6 - 4 - 14 - 47 - 99 -132 -131 1894 - 8 - 19 - 27 - 28 - 25 - 22 - 21 - 28 - 60 -103 -130 -125 1895 - 9 - 21 \#151 ; 26 - 25 - 22 - 20 - 20 - 27 - 56 - 97 -119 -118 1896 - 5 - 16 \#151 ; 20 - 22 \#151 ; 18 - 17 - 19 - 29 - 59 - 95 -118 -114 1897 + 5 - 7 \#151 ; 14 - 15 \#151 ; 15 \#151 ; 17 - 20 - 29 - 54 - 87 -110 -106 1898 - 3 - 12 - 17 - 14 \#151 ; 12 - 15 - 19 - 32 - 58 - 90 -110 -107 1899 - 7 - 14 \#151 ; 16 - 14 \#151 ; 12 \#151 ; 12 - 14 - 23 - 48 - 75 - 95 - 94 1900 + 11 3 - 2 - 1 + 1 \#151 ; 5 - 7 - 16 - 40 - 73 - 99 -100 FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
from Sunspot Minimum Years .
( Unit O'ly .
) 131i .
14k .
15h .
161i .
I71i .
181i .
19h .
20li .
21h .
\#166 ; 22k .
23k .
241i .
Range .
A.l ) .
- 12 + 10 + 32 + 25 + 33 + 29 + 28 + 30 + 24 + 14 + 8 \#151 ; 1 63 19 *3 - 38 6 + 30 + 44 + 49 + 49 + 43 + 42 + 35 + 23 + 18 + 9 98 28 *4 -87 32 + 26 + 62 + 75 + 68 + 63 + 58 + 50 + 39 + 24 + 11 188 40*4 -122 \#151 ; 53 + 3 + 38 + 75 + 92 + 89 + 81 + 70 + 57 + 45 + 31 246 52 2 -115 \#151 ; 49 + 9 + 61 + 94 + 108 + 107 + 98 + 81 + 58 + 45 + 33 273 62 0 - 88 \#151 ; 32 + 10 + 60 + 93 + 108 + 109 + 102 + 79 + 57 + 35 + 19 249 56 -1 - 88 \#151 ; 33 + 28 + 71 + 105 + 121 + 117 + 104 + 83 + 58 + 33 + 13 248 61 *2 - 90 - 32 + 27 + 57 + 76 + 79 + 67 + 58 + 46 + 30 + 15 + 5 205 42 4 - 64 \#151 ; 8 + 34 + 64 + 71 + 64 + 66 + 61 + 52 + 41 + 28 + 8 179 40 *6 - 50 \#151 ; 13 + 26 + 47 + 45 + 46 + 45 + 43 + 40 + 28 + 17 + 7 121 28 *6 - 5 + 26 .
+ 43 + 39 + 38 + 35 + 30 + 23 23 + 9 0 - 9 85 22 2 - 14 0 + 1 9 + 16 + 20 + 20 + 18 + 15 + 17 + 15 + 9 + 5 49 12 *2 - 17 4 .
+ 28 + 31 + 35 + 33 + 30 + 28 + 25 + 15 + 9 + 1 71 20-2 - 81 - 27 + 22 + 53 + 67 + 08 + 66 + 61 + 53 + 41 + 29 + 14 * 179 40 0 - 95 37 + 19 + 62 + 92 + 104 + 100 + 91 + 72 1 + 51 + 32 + 17 241 53 '8 - 64 \#151 ; 18 + 23 + 49 + 64 + 68 + 65 + 60^ + 50 + 36 + 23 + 11 161 37 5 Force , from Individual Years .
( Unit O'ly .
) j 131i .
141i .
15k .
16h .
l7h .
18k .
19h .
20k .
21k .
y 22k .
23k .
24k .
Range .
A.D. - 61 - 15 + 23 .
+ 47 + 61 + 65 + 62 + 56 + 46 + 33 + 22 + 9 152 35 *3 - 78 - 18 + 30 + 70 + 89 + 99 + 93 + 79 + 60 + 38 + 20 + 4 211 48 *5 - 84 - 19 + 40 + 80 + 104 + 115 + 108 + 97 + 76 + 50 + 25 + 2 238 58 4 -101 - 35 + 31 + 72 + 99 + 102 + 94 + 82 !
+ 66 + 44 + 25 + 10 234 52 1 - 91 - 25 + 33 + 81 + 109 + 117 + 108 + 98 + 76 + 50 + 30 + 12 247 59 4 - 82 - 15 + 39 + 77 + 103 + 109 + 103 + 89 + 69 + 42 + 21 + 5 228 54 *7 - 82 - 21 + 33 + 70 + 91 + 97 + 93 + 86 + 71 + 50 + 31 + 13 215 52 9 - 77 - 28 + 21 + 54 + 76 + 87 + 84 + 79 + 68 + 53 + 33 + 19 197 48 3 - 71 - 17 + 30 + 63 + 82 + 91 + 85 + 77 + 63 + 45 J + 27 + 12 201 48 0 - 63 - 17 + 28 + 55 + 73 + 79 + 77 + 69 + 56 + 39 + 21 + 6 174 42 0 - 70 - 23 + 18 + 44 + 59 + 61 + 57 + 53 + 48 + 36 + 27 + 18 161 36 3 vol. coxvi.\#151 ; a. Table XIII.\#151 ; Diurnal Inequality in Total 111 .
2li .
311 .
4li .
5h .
6h .
7h .
81i .
9h .
lOli .
nil .
12h .
; J anuar j . .
\#151 ; 4 7 8 _ 3 + 4 + 8 !
+ 4 \#151 ; 8 - 34 - 57 - 65 - 63 February . .
- 1 - 13 - 16 - 6 0 + 4 + 5 - 5 - 37 - 86 -108 - 92 March " . .
+ 12 + 3 - 2 - 1 + 7 + 10 + 11 - 15 - 82 -153 -195 -176 April . . .
+ 32 + 14 + 6 + 4 + 10 + 17 + 8 - 34 -115 -198 \#151 ; 245 -232 May ... .
+ 29 + 18 + 12 + 12 + 10 - 6 - 46 - 95 -160 -228 -261 -234 June . . . .
+ 15 - 3 - 6 - 2 - 5 33 - 67 - 117 -168 -217 -235 -204 July . . . .
+ 19 - 7 - 18 - 18 - 19 - 48 - 79 - 118 -169 -207 -231 -207 August . . .
+ 29 + 15 + 13 + 8 + 9 - 1 - 32 - 82 -149 -200 -218 -188 September .
+ 28 + 17 + 9 + 7 + 2 4 - 23 - 71 -139 -193 -204 -164 October . . .
+ 16 + 10 + ' 3 + 6 + 9 + 8 + 5 - 19 - 79 -146 \#151 ; 172 -146 November . .
- 7 - 10 - 8 - 3 + 5 + 4 - 4 - 13 - 47 - 90 - 97 - 70 December . .
- 4 - 10 - 7 - 3 + 9 + 16 + 14 + 4 - 19 - 49 - 56 - 57 Year ] ( 11 years ) J \gt ; + 14 + 2 - 2 0 + 3 i - 2 - 17 - 48 -100* -152 .
-174 -153 Sunspot maximum j + 15 + 2 ; - 4 - 2 + 2 2 - 20 - 55 -116 -179 -205 -181 Sunspot 1 minimum / + 13 + 4 + 1 + 4 + 7 + 1 - 12 - 37 - 81 -123 * -139 -121 Table XIV.\#151 ; Diurnal Inequality lh .
2h .
3h .
4h .
5h .
6h .
7h .
8h .
9h .
JOh .
nh .
12h .
January . . .
-0 04 / -0-07 / -0 17 / -0*28 -0 43 -0 53 / -0*51 -0 33 t + 0*03 + 0*32 / + 0 51 / + 0*36 February . .
-0 14 -0-15 -0-17 -0 25 -0 39 -0*51 -0 52 -0 34 + 0*05 + 0*35 + 0 55 + 0*45 Marcli . . .
-0 32 -0 33 -0 36 -0*36 -0*43 -0 46 -0 31 + 0-08 + 0*61 + 0*96 + 1 01 + 0*67 April . . .
-0 48 -0*40 , -0 37 -0-36 -0 33 -0-26 + 0*04 + 0-51 + 1 05 + 1 46 + 1 *39 + 0*89 May . . . .
-0 38 -0 -31 -0 25 -0 17 -0 03 + 0-22 + 0 62 + 1 *00 + 1 19 + 1 16 + 0*89 + 0*49 June . . .
-0-36 -0*31 -0*28 -0*23 -0-06 + 0*27 + 0 70 + 1 07 + 1 *34 + 1 -35 + 1 05 + 0*65 July ... .
-0*38 -0 35 -0-31 -0*25 -0-07 + 0*20 + 0-60 + 1 -02 + 1 *33 + 1 43 + 1*17 + 0*69 August . . .
-0*55 -0 44 -0'41 -0*30 -0 09 + 0-18 + 0-64 + 1*19 + 1 49 + 1 *48 + 1 *11 + 0 *51 September . .
-0 53 -0 47 -0*45 -0 46 -0*43 -0*22 + 0*18 + 0 70 + 1 17 f 1 36 + 1*16 + 0*56 October . .
-0*44 -0-44 -0-49- -0*56 -0 65 -0 60 -0-37 + 0-15 + 0*78 + 1 *18 + 1 *15 + 0*86 November . .
-0*16 -0*20 -0 29 -0-38 \#151 ; 0 ; 50 -0 58 \#151 ; 0 52 -0-25 + 0*23 + 0*55 + 0 67 + 0*53 December . .
+ 0*01 -0*05 -0*16 -0 -27 -0-41 -0 51 -0-48 -0 37 -0*10 + 017 + 0 33 + 0*32 Year \ ( 11 years ) J -0*31 -0*29 -0*31 -0*32 -0*32 -0*23 + 0*01 4 0*37 + 0*76 + 0 98 + 0*92 + 0*58 Sunspot \ maximum J -0*43 -0*40 -0 42 -0-43 -0 39 -0*29 + 0-03 + 0*49 + 0*98 + 1 25 + 1 *20 + 0*82 Sunspot 1 minimum / -0*21 -0 18 -0 19 -0*22 -0 24 -0'17 + 0 01 + 0 29 + 0*60 + 0 76 + 0*66 + 0*36 FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
213 Force , from 11 years .
( Unit 0-ly .
) 13h .
14h .
151i .
leu .
l7h .
18h .
19h .
20h .
21h .
221i .
23h .
24h .
Range .
A.D. - 32 + 10 4 28 4 25 4 32 4 36 4 31 4 37 4 31 4 17 4 10 4 2 102 23 4 - 62 - 14 4 29 4 50 4 59 4 61 4 56 4 57 4 47 4 34 4 24 4 14 169 36 *7 -119 \#151 ; 40 + 37 4 86 4 104 4 105 4 103 4 95 4 82 4 65 4 41 4 22 300 65 *2 -170 \#151 ; 65 4 19 4 74 4 121 4 141 4 145 4 132 4 111 4 91 4 73 4 55 392 88 *3 -163 \#151 ; 69 + 18 4 90 4 153 4 190 4 191 4 170 4 137 4 103 4 79 4 53 452 105 *3 -137 \#151 ; 48 + 33 4 104 4 162 4 193 4201 4 182 4 143 4 100 4 66 4 43 436 103 *5 -146 \#151 ; 52 + 43 4111 4 168 4 202 + 202 4 186 4 150 4 112 4 76 + 48 433 109 *8 -121 - 35 4 44 4 91 4 124 4 140 4134 4 126 4 107 4 84 4 60 4 42 358 85 *5 - 94 - 20 4 38 4 82 4 100 4111 4 120 4 113 4 98 4 83 4 65 + 40 324 76 0 - 92 - 30 4 28 4 61 4 73 4 82 4 82 4 83 4 78 4 62 4 46 4 32 255 57-0 - 38 4 11 + 39 4 53 4 58 4 55 4 51 4 44 4 38 4 21 4 9 - 1 155 32 *3 - 37 - 7 4 11 4 24 4 30 4 31 4 28 4 25 4 24 4 18 4 12 4 4 88 20 *8 -101 - 30 4 31 4 71 4 99 + 113 4 112 4 104 4 87 4 66 4 47 4 30 287 66 *6 -124 - 38 4 35 4 87 4 122 4 138 4 136 4 125 4 103 4 75 4 53 4 33 343 77 *2 - 76 - 19 4 \#187 ; 4 54 4 72 * 4 82 4 83 4 78 4 67 4 52 4 38 4 25 222 50*7 in Inclination , from 11 Years .
131 ) .
141 ) .
15h .
16h .
171 ) .
18h .
19h .
j 20h .
21h .
22 h. 231i .
241 ) .
Range .
A.D. / / / / / / / / / / / / / / + 0 13 4 0*07 4 0*14 4 0*20 4 0*19 40*11 4 0*07 4 0*08 4 0*09 4 0*04 4 0*02 0*00 1 *03 0*196 + 0-27 4 0*17 4 0*21 4 0*27 4 0*26 4 0*18 40 09 -0*01 -0*05 -0*09 -0*12 -0*12 1 *07 0*238 + 0-32 40*13 4 0*07 4 0*09 4 0*13 4 0*02 -0*11 -0*18 -0*24 -0 *30 -0*35 -0*35 1 *47 0 *341 + 0-47 4 0*17 -0*02 -0*17 -0*29 -0*38 -0*44 -0*47 -0*47 -0*49 -0*51 -0 53 1 *99 0*498 + 0 24 4 0*04 -0*14 \#151 ; 0 *31 -0*50 -0*63 -0 11 -0*61 -0*53 -0*47 -0*44 -0*40 1 *90 0*489 + 0-33 4 0*02 -0*26 -0*39 -0*56 -0*71 -0 83 -0*77 -0*64 -0*54 -0*45 -0*39 2*18 0*565 + 0-27 -0*05 -0*29 -0*41 -0*51 -0*61 -0 lit -0*71 -0*62 -0*53 -0*48 -0*41 2*15 0*559 + 0-09 -0*11 -0*18 -0*19 -0*25 -0*44 -0*66 -0 69 -0*67 -0*62 -0*57 -0*56 2*18 0 *559 + 0-17 4 0*05 4 0 *07 4 0*10 4 0*03 -0*17 -0*34 -0*42 -0*45 -0*50 -0 5T -0*55 1 *93 0*463 + 0-33 4 0 *31 " 4 0*23 4 0*23 4 0*09 -0*07 -0*17 -0*22 -0*28 - 0 *37 -0 *43 -0*44 1 *83 0*460 + 0-37 4 0*35 4 0*32 4 0*22 4 0*08 4 0*02 -0*01 -0*04 -0*04 -0 *09 -0*11 -0 *15 1 *25 0*278 + 0 23 4 0*21 4 0*21 + 0*19 + 0*13 40*11 4 0*09 4 0*11 4 0 *10 4 0*08 4 0*04 4 0 *03 0*84 0*196 4 0*28 + 0-11 4 0*03 -0*01 -0*10 -0*21 -0*31 -0 33 -0 *32 * -0*32 -0 33 -0*32 1 *31 0 *336 4 0*46 + 0-21 4 0*06 -0*04 -0*16 -0*31 -0*42 -0*44 -0*44 -0 45 -0 45 -0*43 1 *70 0*458 40*11 0 00 -0*03 -0*02 -0*05 -0*15 -0*22 -0*22 -0*22 -0*22 -0*23 -0*23 1 *00 0*233 2 a 2 Table XV.\#151 ; Diurnal Inequality in lh .
2h .
3h .
4b .
5h .
6h .
7h .
8h .
9h .
10h .
11h .
12h .
/ / , , / , / / / / / / 1890 -0 33 -0*19 \#151 ; 0 *19 -0*22 -0 33 -0*15 + 0*04 + 0*30 + 0*62 + 0 16 + 0*65 + 0*32 1891 -0*30 -0*28 -0*30 -0 34 -0*33 -0*20 + 0*03 + 0-38 + 0*76 + 0 97 + 0*91 + 0 59 1892 \#151 ; 0 '45 -0*41 -0 *41 -0-40 -0 32 -0*23 + 0*06 + 0*51 + 1-00 + 1 39 + 1*20 + 0 77 1893 -0-44 -0*41 -0*41 -0-43 -0-43 -0*29 + 0*05 + 0-54 +1 06 + 1 35 + 1 *31 + 0-88 1894 -0-42 \#151 ; 0 *39 -0 44 -0-45 -0 43 -0-33 o-oo + 0*47 + 0*96 + 1 23 + 1 *20 + 0 90 1895 -0*39 -0-40 -0'40 -0 43 -0*40 -0*30 + 0-01 + 0*43 + 0*89 + 1 10 +1 *09 + 0*72 1896 -0*31 -0*28 -0-30 -0-30 -0 33 -0-24 -0*03 + 0*29 + 0*70 + 0 94 + 0*90 + 0*58 1897 -0 26 -0-24 -0*26 -0-29 -0 30 -0*26 -0*08 + 0*25 + 0*58 + 0-83 + 0*75 + 0-46 1898 -0*26 -G-25 -0*25 -0 27 -0 21 -0*18 + 0*01 + 0*32 + 0*62 + 0 76 + 0*69 + 0*40 1899 -0*26 -0 23 -0*22 -0*24 -0*27 -0*21 -o-oi + 0*30 + 0*65 + 0*83 + 0*71 + 0*42 1900 -0 15 -0 15 -0*17 -0*20 -0*20 -0*16 o-oo + 0*26 + 0*55 + 0 71 + 0*63 + 0*33 Table XVI.\#151 ; Diurnal Inequality in North 111 .
2h .
31i .
4h .
5h .
61i .
7h .
8h .
9h .
10h .
lib. 12h .
January . .
+ 22 + 23 + 31 + 45 + 66 + 78 + 76 + 51 - 6 - 66 \#151 ; 112 -Ill , February . .
+ 44 + 36 + 36 + 48 + 67 + 83 + 84 + 59 - 2 - 74 -135 -141 March . . .
+ 72 + 68 + 70 + 70 + 83 + 89 + 78 + 33 - 60 -158 \#151 ; 227 -222 April . . .
+ 93 + 77 + 73 + 75 + 79 + 82 + 56 - 8 -114 -231 -291 -275 May ... .
+ 79 + 69 + 64 + 62 + 59 + 33 - 22 - 91 -158 -220 -248 -229 June . . . .
+ 69 + 60 + 61 + 68 + 64 + 28 - 32 - 100 -171 -228 -246 -228 July . . . .
+ 74 + 66 + 61 + 64 + 60 + 28 - 30 - 101 -176 -236 -257 -232 August . . .
+ 104 + 89 + 88 + 79 + 65 + 39 \#151 ; 27 - 114 -200 -259 -267 -224 September .
+ 104 + 95 + 92 + 94 + 90 + 67 + 18 - 63 -164 -245 -272 -220 October . .
+ 85 + 81 + 83 + 91 + 103 + 98 + 74 + 13 , - 88 -190 -242 -230 November . .
+ 38 + 37 + 45 + 58 + 75 + 85 + 76 + 43 - 28 -101 -144 -140 December . .
+ 17 + 16 + 27 + 40 + 61 + 70 + 72 + 54 + 14 - 42 - 82 - 98 Winter . . .
+ 30 + 28 + 35 + 48 + 67 + 80 + 77 + 52 - 6 - 71 -118 -123 Equinox . .
+ 89 + 80 + 79 + 82 + 89 + 84 + 56 - 6 -.106 -206 -258 -237 Summer . .
+ 82 + 71 + 68 + 68 + 62 + 32 - 28 - 102 -176 -236 -255 -228 Year 1 ( 11 years ) j + 67 4 60 + 61 + 66 + 73 + 65 + 35 - 19 - 96 -171 -210 -196 Sunspot \ maximum J + 86 + 80 + 81 + 86 + 89 + 81 4 41 - 27 -121 -211 -260 -245 Sunspot 1 minimum j + 48 + 40 + 41 + 48 + 57 + 51 + 29 - 12 - 75 .
i -135 -163 -149 FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
Inclination , from Individual Years .
13h .
14h .
15h .
J till .
I7h .
]8ii .
19h .
201i .
21h .
22li .
23h .
241 ) .
Range .
A.D. / / / / / / / f / / / / / / + 0-06 -0*04 -0-05 -0 00 -0*06 -0*14 -0*20 -0*19 -0*19 -0*18 -0*20 -0 33 0*99 0 229 + 0-27 + 0*08 -0 04 -0*04 -0-14 -0*23 -0*28 -0*29 -0-28 -0-30 -0*32 -0*32 1 *31 0 -333 + 0*44 + 0*19 + 0 04 -0*07 -0*19 -0-32 -0-42 -0 43 -0*45 -0 46 -0 41 -0 41 1*76 0-458 + 0*45 + 0*16 -0*01 -o-ii -0-20 -0 33 -0*44 -0 41 -0*45 -0*46 -0*46 -0-46 1 *82 0-483 + 0*52 + 0 *28 + 0*12 -0*01 -0 16 -0*35 -0*46 -0-48 -0-47 -0-46 -0*44 -0*41 1 *73 0-475 + 0*42 + 0*23 + 0*10 + 0*01 -0*08 -0*24 -0*36 -0*39 -0*40 -0 43 -0 43 -0*39 1 *52 0-417 + 0*29 + 0*14 + 0*05 + 0 01 -0*07 -0*19 -0-29 - 0 33 -0*30 -0*31 -0*31 -0 33 1*26 0-325 + 0*23 + 0*14 + 0*10 + 0*07 -0*05 -0*17 -0*27 -0-27 -0*23 -0*23 -0 26 \#151 ; 0 *23 1 12 0 -283 + 0-18 + 0*07 + 0*06 + 0*06 -0*03 -0*11 -0*24 -0-26 -0*26 -0 26 -0 31 -0 31 1 *03 0-265 + 0*19 + 0*06 + 0*03 + 0*03 -0*04 -0-14 -0*22 -0*24 -0*25 -0*28 -0 31 -0*30 1 13 0-268 + 0*08 -0*04 -0*06 -0 04 -0*05 -0*15 -o n -0 23 -0*20 -0*19 -0*18 -0'16 0*94 0-213 Component , from 11 Years .
( Unit O'ly .
) 13h .
14h .
15h .
16h .
I7h .
j 18h .
19h .
20h .
2 lh .
22 h. 23h .
24h .
Range .
A.D. - 79 - 47 _ 34 \#151 ; 35 \#151 ; 24 - 5 + 9 + 18 + 22 + 27 + 27 + 25 190 43 *3 -119 - 89 - 63 - 43 - 27 - 8 + 10 + 31 + 45 + 51 f 55 + 52 225 58 *4 -179 -123 - 64 - 20 + 6 + 32 + 56 + 69 + 79 + 85 + 84 + 79 316 87 -8 -226 -146 - 66 - 1 + 55 + 95 + 112 + 116 + 114 + 114 + 110 + 107 407 113*2 -187 -123 - 48 + 22 + 90 + 137 + 158 + 143 + 124 + 108 + 97 + 87 406 110 *8 -183 -115 - 37 + 26 + 91 + 139 + 168 + 159 + 132 + 107 + 89 + 79 414 111 *7 -181 -109 - 29 + 35 + 92 + 134 + 156 + 153 + 133 + 112 + 97 + 83 413 112 *5 -163 - 97 - 34 + 16 + 62 + 105 + 133 + 137 + 132 +121 + 110 + 106 404 115 *5 -157 -101 - 56 - 15 + 22 + 60 + 91 + 106 + 109 + 114 + 120 + 111 392 107*7 -183 -124 - 73 \#151 ; 36 0 + 34 + 57 + '73 + 86 + 96 + 99 + 93 345 97 *2 -116 - 87 - 58 - 31 - 2 + 13 + 25 + 35 + 42 + 46 + 45 + 44 229 58 *9 - 86 - 67 50 \#151 ; 34 \#151 ; 15 - 4 + 6 + 10 + 18 + 22 + 24 + 22 174 39 *9 -100 - 73 \#151 ; 51 \#151 ; 36 17 - 1 + 13 + 23 + 32 + 37 + 38 + 36 203 49 *7 -186 -123 - 65 - 18 + 21 + 55 + 79 + 91 + 97 + 102 + 103 + 98 361 100 *4 \#151 ; 179 -111 \#151 ; 37 + 25 + 84 + 129 + 154 + 148 + 130 + 112 + 98 + 89 409 112 *7 -155 -102 51 - 10 + 29 + 61 + 82 + 87 + 86 + 84 + 80 + 74 297 84 *2 -198 -134 67 - 11 + 38 + 78 + 102 + 109 + 108 + 104 + 99 } 93 369 106 *2 -111 - 69 1 30 - 2 + 21 + 45 + 61 + 64 + 64 + 62 + 60 + 55 227 62 *2 Table XVEI.\#151 ; .Diurnal Inequality in West lh .
2li .
3h .
41i .
51i .
6b .
7h .
8h .
9h .
10h .
11h .
12h .
J anuary . .
- 62 - 49 - 35 - 28 - 20 - 15 - 18 - 32 - 37 0 + 55 + 120 February . .
- 83 - 70 - 63 \#151 ; 55 - 48 - 42 - 38 - 52 - 66 - 26 + 57 + 141 March . . .
- 75 - 73 - 73 - 74 - 72 - 72 -102 -158 \#151 ; 173 -102 + 42 + 197 April . . .
- 50 - 58 - 68 - 84 - 98 -127 -186 \#151 ; 338 -232 -140 + 23 + 204 May ... .
- 52 - 67 - 84 -113 -163 -211 -356 -251 -208 - 84 + 76 + 221 June . . . .
- 47 - 64 - 86 -124 -187 -248 -284 -388 -242 -126 + 35 + 186 July . . . .
- 48 - 70 - 89 -121 -186 -238 -263 -365 -226 -124 + 26 + 182 August . . .
- 60 - 76 %- 91 -114 -158 -204 -243 -349 -190 - 62 + 100 + 257 September . .
- 72 - 78 - 88 - 96 -103 -127 -165 -190 -154 - 43 + 108 + 252 October . .
- 64 - 58 - 52 - 44 - 41 - 46 - 73 -130 -154 - 86 + 61 + 184 November . .
- 61 - 44 - 31 - 25 - 21 - 21 - 24 - 49 - 71 - 31 + 53 + 134 December . .
- 60 - 44 - 28 - 14 - 7 - 2 - 5 - 13 - 26 - 2 + 48 + 104 Winter . . .
- 67 - 52 - 39 - 31 - 24 - 20 - 21 - 37 - 50 - 15 + 53 + 125 Equinox . .
- 65 - 67 - 70 - 75 - 79 - 93 -132 \#151 ; 179 -178 - 93 + 59 + 209 Summer . .
- 52 - 69 - 88 -118 -174 -225 -262 -363 -217 - 99 + 59 + 212 Year 1 ( 11 years ) J - 61 - 63 - 66 - 74 - 92 -113 -138 -160 -148 - 69 + 57 + 182 Sunspot 1 maximum J - 76 - 81 - 85 - 94 -114 -138 -167 -193 -181 - 90 + 54 + 201 Sunspot \ minimum j - 45 \#151 ; 44 - 47 - 57 - 73 - 91 -114 -133 -121 - 51 + 58 + 164 FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
Component , from 11 Years .
( Unit O'ly .
) 13h .
141i .
15h . !
161i .
I7h .
1 181i .
19h .
20h .
21h .
22U .
231i .
24h .
Kange .
A.D. +151 + 134 + 87 + 58 + 37 + 15 \#151 ; 13 \#151 ; 43 \#151 ; *\lt ; r o - 84 \#151 ; 83 _ 75 241 55 *3 + 190 + 195 + 148 + 88 + 54 + 28 + i i - 27 - 60 - 81 - 92 94 289 75 0 + 282 + 288 + 226 + 134 + 57 + 23 2 - 21 - 45 - 69 ~ 67 - 76 461 103 *9 + 305 + 311 + 246 + 170 + 101 + 47 + 16 - 3 - 18 - 35 - 41 - 45 549 118 6 + 287 + 288 + 239 + 178 + 125 + 77 + 46 + 23 + 6 - 9 - 23 - 37 544 130 *2 + 267 + 299 + 275 + 220 + 156 + 109 + 75 + 55 + 32 + 16 - 6 - 27 587 143 *9 + 274 + 303 + 272 + 207 + 142 + 94 + 69 + 53 + 34 + 14 - 9 - 29 568 139 1 + 328 + 315 + 244 + 147 + 70 + 34 + 29 + 18 + 4 - 17 33 - 48 577 128 *8 + 309 + 277 + 195 + 108 + 48 + 23 + 8 - 13 - 31 - 44 57 - 68 499 110 -7 + 240 + 226 + 171 + 97 + 53 + 26 - 3 - 33 58 - 7J - 75 - 70 394 88 *2 + 105 + 144 + 101 + 72 + 46 + 19 !
" 8 - 39 - 69 - 82 - 84 77 249 61 *3 + 133 + 115 + 83 + 58 + 33 + 6 - 22 48 71 - 8 * \#151 ; 81 77 215 48 *4 + 161 + 147 + 105 + 69 + 43 + 17 .
- 10 39 _ 67 - 82 \#151 ; 85 \#151 ; 81 246 60 -0 + 284 + 276 + 210 + 127 + 65 + 30 + 5 18 38 - 53 - 60 65 463 105 *4 + 289 + 301 + 258 + 188 + 123 + 78 + 55 + 37 j + 19 + 1 \#151 ; 18 35 564 135 0 + 245 + 241 + 191 + 128 + 77 + 42 + 16 - 7 - 29 - 45 54 - 60 405 98 *2 + 280 + 288 + 236 + 167 + 103 + 60 + 29 - 1 - 25 - 44 - 59 - 70 481 118 *2 + 213 + 198 + 144 + 88 + 46 + 24 + 8 - 10 - 27 - 39 - 45 - 46 346 78 -6 As fig. 2 shows , the difference between sunspot maximum and minimum is mainly a matter of amplitude ; but in winter , in particular , the tendency in the afternoon curves to a plateau\#151 ; i.e. , the maintenance of a uniform value\#151 ; is especially characteristic of sunspot minimum .
Active regular changes in H at midwinter are mainly confined to the hours 2 a.m. to 2 p.m. Whether we consider the range or the A.D. , December and January are clearly the months in which the diurnal inequality is least .
In the case of the eleven years and the sunspot maximum group of years , the minimum is in December , but in the S minimum group it falls in January .
The differences , however , between the two months are not so decisive as to justify the conclusion that what is true of this particular eleven years is invariably true .
The sunspot minimum group in particular , it should be remembered , contained only three years .
The maximum amplitude appears in June and July ; but May and August are not far behind , especially in the sunspot minimum group of years .
In sunspot maximum years April closely resembles May as regards range and A.D. , but it exhibits a decided tendency to the evening plateau and morning minimum characteristic of winter .
In sunspot minimum years April falls slightly behind September as regards amplitude , while September in all years falls markedly short of August and exhibits a distinct trace of the winter characteristics .
These features , in conjunction with the desirability of having the same number of months in each season , seem to justify the division of months adopted .
Any grouping which combined February with March , or October with November , seems less appropriate .
One or two features in Table VIII .
call for remark .
Whether we take the range or the A.D. , the amplitude of the mean diurnal inequality for the year was greatest in 1893 , the year of largest sunspot frequency , and least in 1890 and 1900 , the years of least sunspot frequency .
As will be seen presently , the inequality range and the sunspot frequency show a very close parallelism in their variation .
The changes in the type of the inequality from year to year are small , and their elucidation requires a more sensitive method , such as the analysis into Fourier waves presently discussed .
A tendency can , however , be recognised for the hour of the forenoon minimum to be slightly earlier when sunspots are few than when they are numerous .
S 8 .
The V inequality data for the 12 months in Table IX .
are shown graphically in fig. 3 , while fig. 4 contrasts the diurnal inequalities for the seasons and the year derived from the whole eleven years and from the sunspot maximum and minimum groups of years , as given in Tables IX .
, X. , and XI .
The V inequality data are not quite so smooth as those for H , especially in winter .
The fact that the H curves were smoothed , while the V curves wei'e not , presumably partly accounts for this .
The variation in the type of the inequality throughout the year is less in V than in H. In some months two distinct maxima and minima are visible in V , but the principal maximum always occurs in the afternoon between 4 p.m. and 7 p.m. , while the principal minimum , in the forenoon , is as uniform in its time of incidence as the corresponding minimum in H. The time intervening between the morning minimum FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
and afternoon maximum is less for V than for H , and , speaking generally , the preponderance in changes by day over those by night is greater in V. A further notable difference is that the season in which the double maximum and minimum are most in evidence is summer for V , but winter for H. Also the secondary maximum in the forenoon is in V most prominent at sunspot maximum , while in H it is most prominent at sunspot minimum .
In winter the changes during the night hours are exceedingly slow in V as in H ; but while the fall to the forenoon minimum is the conspicuous feature in H , it is the rise after the morning minimum that stands out in V. The minimum amplitude occurs as distinctly in December and January for V as for H. In V the lower value is found in December , both in sunspot maximum and minimum , especially the latter , again a difference from what we found in H. The smallness of the amplitude in V in December at sunspot minimum was , however , in considerable measure due to one year , 1890 .
During the last three months of that year the changes apparent in the V curves , regular and irregular alike , became so small that some defect in the magnetograph was suspected at the time , and the V data in the tables published in the Annual Report of the Kew Committee were confined to the first nine months of the year .
There was , however , nothing but the smallness of the movements to suggest instrumental defect , and as the movements rapidly increased in 1891 , without anything being done to the instrument , and the scale value determination showed nothing abnormal , the phenomenon was presumably real , and not of instrumental origin .
As regards the month when the V inequality is largest , the decision depends upon what we accept as the criterion of amplitude .
May comes distinctly first if we take the range , but if we take the A.D. July comes first in the case of the eleven years , and still more so in the case of the sunspot maximum group of years .
August falls markedly short of the other summer months , and is inferior even to April .
Table XII .
shows that in V , as in H , 1890 and 1900 were the years when the amplitude of the inequality was least ; 1893 , however , the year of sunspot maximum , was not the year of largest amplitude in V , being exceeded by 1892 and 1894 , which were also much more disturbed .
The difference in amplitude between years of many and few sunspots is less conspicuous in Table XII .
than in Table VIII .
, and , as will be seen more clearly later , the parallelism between the amplitude of the inequality range and the development of sunspots was not so close in V as in H. Whether this last is a real natural phenomenon , independent of the lesser reliability of the V data in respect especially of temperature correction , is perhaps open to some doubt .
S 9 .
If we write the first of equations ( 2 ) , p. 198 , in the form T AT = H AH + V AV , and remember that at Kew V = 2'4H roughly , while the range of the diurnal inequality in V is about three-fourths that in H , we see at once that so far as diurnal changes are concerned , the influence of V on T very considerably exceeds that of H. VOL. CCXVI.\#151 ; A. 2 H iO Avo/ l 14 16 JS ZC Z2 Juno Fig. 5 .
Total force .
FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
Also the diurnal inequalities in V and H follow a fairly similar course , thus we know a priori that the diurnal inequality in T must in general resemble somewhat closely that in V. In winter , however , the amplitude of the Y inequality suffers a considerably larger relative reduction than that of H , and the difference of type between the inequalities in the two elements is then at its maximum .
Thus we may expect the resemblance of the T to the Y inequality to be less close in winter , and the influence of the H inequality to be then more apparent .
The accuracy of these anticipations is readily recognised if we compare Table XIII .
and fig. 5 with Table IX .
and fig. 3 .
Owing to the difference in type between the Y and H inequalities in December and January , their contributions to the T inequality tend sensibly to neutralise one another .
Thus the amplitude of the T inequality at midwinter is markedly less than that of H , while in June and July the amplitudes in the two elements are very similar .
Consequently the variation in the range in the course of the year is decidedly more conspicuous in Table XIII .
than in Tables Y. or IX .
A conspicuous feature in Table XIII .
is the uniformity in the time of occurrence of the principal minimum .
It is shown at 11 a.m. in every month except December , when the 11 o'clock value just exceeds that for noon .
The existence of a second minimum in the early morning is recognisable in most of the monthly curves of fig. 5 ; there is at least an arrest in the rate of change of the element .
The maximum , or principal maximum , always occurs in the afternoon , usually from 6 to 7 p.m. Amongst the curves of fig. 7 are three contrasting the mean diurnal inequality in T for the year in the years of sunspot maximum and minimum and the average year .
The variation of type with sunspot frequency makes little appeal to the eye .
S 10 .
From the second of formulae ( 2 ) or ( 3 ) connecting the I inequality with those for Y and H , it appears that the contributions from Y and H oppose one another , but that the H contribution must be largely dominant .
The phases in the I and H inequalities are opposite , so a comparison of fig. 6 , showing the inequalities in I , with fig. 1 does not at first sight show the dominance of H. A little consideration will , however , make this clear , if we remember the difference of phase .
Answering to the prominent forenoon minimum in H , we have a prominent forenoon maximum in I , its time of occurrence being visibly earlier in summer than in winter .
Again , as in H , there is a distinct difference of type between summer and winter .
In summer , the principal or only minimum occurs in the afternoon , just as the principal maximum does in H ; whereas in winter the principal minimum , like the principal maximum in H , is found in the morning .
In most months between 2 and 6 p.m. there is at least an arrest in the fall of I , just as there was an arrest in the rise of H. The opposition in phase between the I and T inequalities is readily recognised in fig. 7 , where the curves from the two elements representing the mean diurnal variation for the year in the average year and in years of sunspot maximum and minimum are juxtaposed .
\#166 ; 2 H 2 0 2 4 6 8 ip Neon 14 16 18 20 22 Z4Alanc/ i Noon in Mic/ il ?
Fig. 6 .
Inclination .
0 2 * 6 8 / oJVbo/ t 14 16 18 20 22 Sunspot moor " Nbrtfv component Sunspot minr .
West component Sunspot Fig. 7 .
Mean diurnal inequalities for the year .
The arrest in the afternoon fall of I , it will be seen , is more prominent at sunspot minimum than at sunspot maximum .
According to Table XIV .
, the December diurnal inequality in I has a decidedly smaller range than the January inequality , but the values of the A.D. for the two months are equal .
In this case the August diurnal inequality markedly exceeds that for May , and is practically equal in amplitude to the June and July inequalities .
The diminished amplitude of the May inequality in I is associated with the corresponding enhanced amplitude of the inequality in V. So far as amplitude is concerned , one would naturally group May with April , September and October .
In amplitude March shows a closer approach to February than to April ; in type it stands about equidistant from the two adjacent months .
S11 .
The graphical presentation of N and W diurnal inequalities , as given in Tables XVI .
and XVII .
, is limited to the mean diurnal inequalities for the year from the eleven years and the groups of years of sunspot maximum and minimum .
These are shown in the six uppermost curves of fig. 7 .
If we take the mean diurnal inequality for the year , the range in H is only about three-fifths that in D , while cos D is three times sin D. It follows from the last of equations ( 2 ) that the influence of the D diurnal changes on W will largely exceed that of the H changes .
Hence the diurnal inequality in W must show a fairly close approach in type to that in D. In the case of N the H diurnal changes exert more influence than the D changes , but the preponderance is less .
As a matter of fact , however , there is a very close resemblance between the diurnal inequalities in H and N. The inequality in D if reversed has a considerable resemblance to that in H , and the contributions from H and D to the N diurnal inequality have opposite signs , so numerically they assist one another .
It is pretty obvious , comparing figs. 1 to 7 , that if we measured east-west changes positive to the east , instead of to the west , and I changes positive from instead of towards the vertical , the diurnal inequality curves for the year in all the six elements considered here , and in D as welfj would agree in having as one of their most prominent features a principal minimum occurring within an hour or two of noon .
According to Table XVI .
, the diurnal inequality in N is least in December , January coming next .
June and July show the largest ranges , but there is little variation in that respect from April to September .
August shows the largest A.D. , April coming next .
The six months April to September agree in having the maximum or principal maximum in the afternoon , while the other six months have it between 5 and 7 a.m. ' Thus the phenomena in N rather favour the sub-division of the year into two seasons , each of six months .
In March and October , however , while the forenoon maximum is the larger , it but very slightly exceeds the afternoon maximum , and the grouping of these months with the four midwinter months would be unsatisfactory .
FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
According to Table XVII .
, December has distinctly the smallest amplitude in the diurnal inequality in W , January coming next .
June shows the largest range and A.D. , July coming decidedly next so far as A.D. is concerned , though slightly inferior to August as regards range .
April closely resembles the midsummer months so far as range is concerned , but fall distinctly short of them in A.D. In this element the most conspicuous difference of type is between the four midwinter months , when the principal minimum occurs between 10 p.m. and midnight , and the other eight months when the principal minimum occurs between 7 and 9 a.m. The maximum , or principal maximum , is the pre-eminent featui'e .
Like the corresponding maximum of westerly declination , it occurs between 1 and 2 p.m. the whole year round .
Comparing Tables XYI .
and XVII .
, we see that the range of the mean diurnal inequality for the year in N is roughly three-fourths that in W ; in winter and equinox , however , the fraction is distinctly larger .
There is less difference between the values of A.D. in the two elements , especially in equinox .
S 12 .
Figs. 8 , 9 and 10 give the vector diagrams for the diurnal inequality forces in the horizontal plane .
The arms of the cross are oriented in and perpendicular to the geographical meridian .
The vector drawn from the centre of the cross , which serves as origin , to the points 1 to 24 , represents in magnitude and direction the value at the hour specified of the horizontal component of the force to which the departure of the magnetic field from its mean value for the day may be ascribed .
The scale cf force is such that each arm of the cross represents 10y .
The vector travels completely round in the general clockwise direction in the course of the 24 hours .
The motion is always direct between 6 a.m. and 10 p.m. , but in the six months , October to March , there is a reversal of direction during some of the night hours .
Details of this reversal are not clearly visible on the scale adopted in fig. 8 , so they are shown on a more open scale in fig. 9 , which is limited to part of the day .
The monthly diagrams in fig. 8 bring home to the eye the great variability in the type of the regular diurnal variation throughout the year .
If the June and December diagrams were alone presented , it would be difficult to believe that they referred to the same physical entity .
There are , however , some features common to all the diagrams .
The time when the vector is directed due south falls in all cases between 10 and 11 a.m. , and so is well in advance of the sun 's crossing of the meridian .
Again the vector in all cases has its maximum value near 1 p.m. The minimum value is usually seen near 6 p.m. , but from April to July it occurs shortly after midnight .
The May , June , and July diagrams closely resemble one another .
The August diagram differs from them in having a distinct bay in the afternoon hours , resembling in that respect the September and March diagrams .
The diagrams for the midwinter months , November to February , have a pretty close family resemblance .
If we take the months in their order , we can recognise a continuous development from January to May , and from July to December ; but it is obvious that in some months , especially Fig. 8 .
Vector diagrams in horizontal plane ( 11 years ) .
Arms of cross each lOy .
FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
in equinox , the amplitude and type at the beginning and end of the month must differ somewhat largely .
The phenomena , for instance , near midnight must be considerably different in the beginning and end of March .
What the March diagrams in figs. 8 and 9 show is a blend of different types .
We want shorter periods than the calendar month if we are to trace the changes minutely .
Fig. 9 .
Vector diagrams in horizontal plane .
Details of variation near midnight ( 24h ) .
( ON = O = 10y .
) Fig. 10 shows the horizontal plane vector diagram for the year in the case of the whole eleven years , as well as in the cases of the sunspot maximum and minimum groups of years .
The diagrams for the sunspot maximum and minimum groups of years are drawn from a common origin , as serving best to bring out points of agreement and difference .
Except as regards the amplitude of the vector , the most noticeable difference between the sunspot maximum and minimum diagrams is that the indentation extending from about 11 p.m. to 5 a.m. is much more prominent in the latter .
This indentation represents an approach to retrograde movement , a characteristic , as we have seen , of the winter season .
Also , while a bay is recognisable in the afternoon hours in all three curves , it is deeper in the sunspot minimum diagram than in the other two .
VOL. ccxvi.\#151 ; A. 2 i The eye can recognise in fig. 10 that the sunspot minimum vector is in advance of the sunspot maximum vector from 10 a.m. to 10 p.m. , the angular difference being considerable in the early afternoon .
The vector has , however , its maximum near I p.m. in both cases .
M SCALE Arms of cross o.M.Mcumelic meriaui/ h Fig. 10 .
Vector diagram in horizontal plane .
Year .
linear mean Sunspot maximum outer carves .
Sunspot mtmnuirns mixer curve .
The year diagram from the eleven years does not show a close resemblance to the diagram of any individual month in fig. 8 , a suggestive fact in connection with the significance of mean diurnal inequalities for the whole year .
The equinoctial months are those whose diagrams show most resemblance to that for the whole year .
FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
S 13 .
Fig. 11 shows vector diagrams in two vertical planes .
The upper six curves refer to the vertical plane through the geographical west , the lower six to the vertical plane through geographical north .
The upper three WY curves and the upper three NY curves represent the mean diurnal inequalities for the three seasons derived from the whole eleven years .
The three lower diagrams in each case represent the mean diurnal inequality for the year , derived respectively from the sunspot minimum group of years , the whole eleven years and the sunspot maximum group of years .
The origin of co-ordinates is marked in each case by a small cross .
The two large crosses show the orientation , one for the WY diagrams , the other for the NY diagrams ; the arms of these crosses each represent lOy .
It will be noticed that Y is measured positively downwards in both cases .
The three WY diagrams for the year differ markedly in amplitude , but comparatively little in type , closely resembling the equinoctial diagram from the eleven years .
The special feature is the beak occurring about 8 a.m. This beak is also prominent in the summer diagram , occurring , however , nearly an hour earlier .
Associated with the beak is an indentation or bay , extending from about 6 p.m. to 7 or 8 a.m. In the summer diagram the curvature is small near the deepest part of the bay , and except at the beak is large only from 6 to 8 p.m. In the diagrams for equinox and the year there is , as it were , a pushing forward of the shore line of the bay from 6 p.m. to midnight , leading to a marked indentation in the early morning hours .
This advance of the shore line of the bay in the afternoon hours has so far developed in the winter diagram that all that remains of the bay is the very deep narrow indentation in the early morning hours .
The beak is still represented in the winter diagram at 9 a.m. , i.e. , about an hour later than in the equinoctial diagram , and two hours later than in the summer diagram , but it is now overshadowed by the protuberance near midnight .
The three NV diagrams for the whole year again differ mainly in amplitude .
They all show a distinct loop near mid-day , somewhat more developed relatively in the sunspot minimum diagram than in the other two .
The sunspot maximum and eleven-year diagrams show also a Very small loop near 4 a.m. In the sunspot minimum diagram this is represented by a sort of tail , with its tip at 5 a.m. The hour marks for several hours on end are so nearly on a straight line , that it is impossible to say whether there is a true loop , as in the eleven-year diagram , or a very narrow indentation .
The equinoctial diagram very closely resembles that for the eleven years .
The summer diagram shows a considerably larger development of the mid-day loop , but the early morning loop has disappeared and is represented by a small tooth .
The winter diagram shows development in an opposite direction .
The mid-day loop has disappeared , and instead of a loop in the early morning there is a long narrow promontory , which like the rest of the diagram is described anticlockwise .
A feature in NV diagrams to which attention was called some years ago by NY vector diagrams .
FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
DR. C. CHREE : DISCUSSION OF KEW MAGNETIC DATA Mr. R. B. Sangster* is the approach which the diagram makes in the early afternoon hours to a straight line .
This feature is prominent in the summer diagram between noon and 4 or 5 p.m. It is also fairly shown in the equinoctial diagram and in the diagrams for the year , especially that from sunspot maximum , but hardly in the winter diagram .
S 14 .
Fig. 12 shows ordinary less quiet day difference curves , i.e. , the ordinate represents the algebraic excess of the ordinary day over the corresponding quiet day value in the diurnal inequality .
All the curves refer to the mean diurnal inequality for the year .
In the case of V , T and I there is only one curve , which corresponds to the whole eleven years .
In the case of H , N and W there are three curves , representing the sunspot maximum and minimum groups of years in addition to the whole eleven years .
The results were not smooth enough to justify drawing curves of continuous curvature .
The scale , which is the same for all the force curves , is much more open than in the corresponding ordinary day curves .
In each case the ordinary less quiet day difference curve shows a considerable resemblance in type to the corresponding disturbed less quiet day difference curve given in a previous paper , but its amplitude is very much less .
In the case of the H difference curves in fig. 12 the amplitude is so small that a longer period than eleven years would have been needed to bring out the character fully during the night hours .
There is a distinct minimum in the curve near noon\#151 ; signifying that the principal daily minimum is more developed in the ordinary than in the quiet day curve\#151 ; but that is perhaps the only unmistakable feature .
The irregularities in the sunspot maximum and minimum curves are too great to warrant deductions as to differences between them .
The N , and still more the W , difference curves have a decidedly larger amplitude and are less irregular , especially the 11-year curves .
The N difference curve has its mid-day minimum well developed , like the ordinary N curve , but it has a relatively better developed night maximum , and there seems to be no secondary maximum and minimum .
The W difference curve differs markedly from both the ordinary and quiet day curves .
There is no morning minimum near 8 a.m. , as in the ordinary day curve , and instead of a sharply defined maximum in the early afternoon , there is a wide plateau for some four hours on either side of noon .
The largest ordinate appears at 4 p.m. , i.e. , from two to three hours later than in the ordinary day curve .
Whether the saddle near 1 p.m. , which is more apparent in the sunspot maximum and minimum than in the 11-year curves , is a real or an accidental feature is open to some doubt .
The Y and T difference curves closely resemble one another , as might have been inferred from the fact that the difference between ordinary and quiet day inequalities is much larger for V than for H. The two curves , while quite unlike the ordinary day Y and T curves , closely resemble the disturbed less quiet day V difference , * ' Roy .
Soc. Proc. , ' A , vol. 83 , p. 428 FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
O 2 4 6 8 to Noon 16 18 20 22 24 Sunspot / / years \gt ; Simspot max ?
\gt ; Sunspot -\#151 ; Sunspot max'7 \\gt ; // Yeans \^pSiuispot Fig. 12 .
Difference curves , ordinary less quiet days .
'234 DE .
C. CHEEE : DISCUSSION OF KEW MAGNETIC DATA curves .
They show one well marked maximum near 6 p.m. , and an equally well marked minimum about 4 a.m. The I difference curve in fig. 12 is also markedly unlike the ordinary day I curve .
It has a considerable resemblance to the W difference curve , but corresponding points occur some hours later in the day in the I curve .
Some doubt must be entertained as to the reality of the depression shown about 4 p.m. The difference curves in fig. 12 indicate on the whole an increased amplitude in the ordinary day inequality as compared with that of the quiet day , but they also indicate a distinct difference in type .
The nature of the difference is most conveniently considered in connection with the harmonic analysis of the inequalities .
One outstanding feature has already been discussed.* Annual Inequality .
S 15 .
Let M represent the mean value of any element for the whole year , and AM the algebraic increment in the twelve months due to secular change , then if the secular change took place uniformly throughout the year the mean value of the element during the nth month should be M+(w\#151 ; I\#151 ; 6 ) AM/ 12 , where n represents 1 , 2 , 3 up to 12 .
This neglects the difference in length between different months .
If the actual mean value for the nth month proves to be pn , the values of pn- { M+ ( n-J-6 ) AM/ 12 } for the 12 months constitute the annual inequality .
If a magnetograph had a known and invariable base line value , the annual inequality of the corresponding element could be determined from curve readings alone .
On the other hand , if an absolute instrument were invariably read when the corresponding element had its mean value for the day , the annual inequality could be got without any reference to the curves .
But in practice we can tell the base line value of the curve only by reference to the absolute observations , and we cannot tell when the element had its mean value without recourse to the curves .
When base values have been assigned to all the curves , a mean value can be found for each day , representing the mean of the hourly values .
From these daily means a monthly mean can be derived , representing , according to the practice of the particular observatory , the mean of all the days of the month , or the mean of all but highly disturbed days , or the mean of a limited number of selected days , e.g. , the 5 international quiet days .
Unless all days are used , the mean of the days employed may * 'Koy .
Soc. Proc. , ' A , vol. 91 , p. 370 .
FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
not fall at the centre of the month , and a question may arise as to whether some allowance should be made for this .
The desirability of selecting the 5 quiet days so that their mean shall fall near the middle of the month is recognised , but such a choice is not always possible .
The practice in vogue at Kew from 1890 to 1900 was to treat the base value of each element as invariable throughout each calendar month , except in so far as it was influenced in the case of H and V by departures of the temperature of the magnetograph magnets from the mean temperature of the month .
Using these base values , it would have been possible to deduce mean values depending on all the days , or on all but the highly disturbed days , in the usual way .
Having regard to the special circumstances , however , a different method\#151 ; presently to be described\#151 ; has been adopted as equally satisfactory and much simpler .
It recognises that it is really on the absolute instruments that the accuracy of the annual inequality ultimately depends .
The magnetographs serve only to show what allowance has to be made to any individual absolute observation to bring it to what it would have been if taken at an hour when the element in question has its mean value for the day .
To take a simple example , suppose we observed the declination daily at noon , and that in the diurnal inequality for a particular month the entry under noon is + 3''0 .
Then to get the mean value of D for that particular month all we have to do is to subtract 3/ -0 from the arithmetic mean of the absolute observations .
The correction \#151 ; 3'*0 applied to individual noon readings will not in general give exactly the mean value for the day , but so far as the monthly mean is concerned that is immaterial .
If we observe not daily but weekly , the 4 or 5 monthly days of observation at noon may all happen to depart from the average day of the month in the same direction , so that the method if applied to the observations of a single year could hardly claim to be satisfactory .
If , however , we deal with the observations not of one but of a number of years , omitting days of large disturbance , accidental features must largely disappear , and this is the course that has been adopted .
The ordinary absolute observation of H consists of two parts , the vibration experiment and the deflection experiment .
The mean times of these two experiments were found , and departures at each of them from the mean value for the day were deduced from the diurnal inequality , and the arithmetic mean of the two was applied with appropriate sign as a correction to the observed value of H. The inequalities , it will be remembered , were calculated for each individual month for H ( and V ) .
When the curves were considerably disturbed at the time of an absolute observation\#151 ; not , of course , a frequent event\#151 ; the observation was simply omitted .
As highly disturbed days had also been omitted from the ordinary day diurnal inequalities , we may reasonably regard these inequalities as appropriate for the correction of the absolute observations .
It had not been the practice to commence the H observation at a fixed hour , and the time required for the deflection experiment varies somewhat according as VOL. CCXVI.--A .
2 K 23G DR. C. CHREE : DISCUSSION OF KEW MAGNETIC DATA observing conditions are more or less favourable .
Still in most months the mean times of the vibration and deflection experiments fell so nearly at fixed hours that one might have applied the same correction to each observation , basing it on a mean time from all the observations .
As a matter of fact , however , this simplification was not adopted when dealing with the 11 years 1890 to 1900 .
After 1900 , as already explained , magnetic conditions were less favourable at Kew .
Still electric tram disturbances are , on the whole , of a nature calculated to impair the accuracy of individual observations rather than that of the arithmetic mean of a number of observations .
It thus appeared worth while investigating the annual inequality deducible from absolute observations made in years subsequent to 1900 .
As corresponding diurnal inequalities from ordinary days were not available , the corrections applied to the absolute observation results to bring them to the mean value for the day were based on the inequalities of the years 1890 to 1900 .
In the case of H , the years 1901 , 1902 , 1903 , 1910 , 1911 , 1912 , and 1913 , having a mean sunspot frequency of 8'8 , had corrections applied from Table YII .
, which is based on the years 1890 , 1899 , and 1900 , with a mean sunspot frequency of 9'6 .
The years 1905 , 1906 , and 1907 , having a mean sunspot frequency of 59'9 , had corrections applied from .
Table VI .
, which is based on the years 1892 to 1895 , with a mean sunspot frequency of 75'0 .
The two years 1904 and 1909 , having sunspot frequencies of 42'0 and 43'9 respectively , had corrections applied from Table V. for the 11 years 1890 to 1900 , with a mean sunspot frequency of 417 .
The year 1908 was omitted , because the deflection distances were increased from two to three at midsummer , and this possibly might have introduced some discontinuity .
The final outcome was that the twelve years dealt with had corrections applied as if their mean sunspot frequency were 31 , whereas it was really 27 .
As we shall see later , the amplitude of the diurnal inequality in H and the sunspot frequency are connected , at least approximately , by a linear relationship .
Extrapolation from one period of years to another is of course always a matter of some uncertainty ; still there is considerable ground for believing that the accuracy of the inequality derived from the second period of years is not greatly inferior to that derived from the first . .
In the case of I the observation had been taken almost invariably in the afternoon , the mean time of observation falling within 30 minutes of 3 p.m. Near this hour I changes slowly , and its departure from the mean value for the day is not large .
To have got out diurnal inequalities of I for each month of the 11 years would have entailed an immense amount of labour .
Thus corrections to the observed values were simply derived from the mean inequalities from the 11 years given in Table XIV .
, and the corrections were calculated for the mean time of all the observations of the month , not for the times of the individual observations , unless these times varied more than was generally the case .
The I observational data from the 14 years 1901 to 1914 were similarly treated , use being again made of the diurnal inequalities in Table XIV .
FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
S 16 Values having been obtained from the absolute observations of each month corrected for diurnal variation , the means for all the months of the same name were summed and meaned , the 11-year period and the later period being treated separately .
The monthly means thus obtained represented the average annual change from mid-January to mid-December , comprising the annual inequality and the secular change .
It remained to eliminate the secular change .
In the case of the 11-year period , the mean annual secular change of H between 1890'5 and 1900'5 derived from the quiet days was +25'9y .
A small difference has been observed at several stations between mean annual values of H derived respectively from quiet days and from all or all ordinary days , the former mean exceeding the latter .
The difference , however , is only of the order 3y , and as 1890 and 1900 were both quiet years of very similar character , any uncertainty of this kind spread over a 10-year period must have been negligible .
A confirmation of the accuracy of the quiet day estimate was derived by taking arithmetic means for 1890 and 1900 of the twelve monthly means of the absolute observations , corrected for diurnal variation .
The mean thus obtained for 1900 exceeded that obtained for 1890 by 259'5y , giving + 2o'95 for the mean secular change .
In the case of the second group of years , the mean secular change in H\#151 ; as obtained from the annually published quiet day results , allowing for changes of constants and procedure\#151 ; was only +5'8^ .
In the case of I the mean secular changes accepted were \#151 ; 2/ '25s for the first period and \#151 ; 1''055 for the second .
The substitution of these respective values for AM in the formula in S 15 led to the annual inequalities for H and I given in Table XVIII .
The annual inequalities given for V in that table were calculated from the formula expressing changes in V in terms of changes in H and I , employing mean values for the numerical coefficients of AH and Al .
At the foot means are given for the ordinary three seasons .
The centre of each season falls at the middle of the year , so these seasonal values are unaffected by the secular change , or by any error that may have been made in estimating its amount .
The greater or less smoothness of the inequalities , and the amount of accordance between the results from the two periods , are the chief criteria for estimating the reliability of the data in Table XVIII .
Both criteria are less favourable towards the inequality in V than towards those in H and I. The V inequality from the 11-year period is very irregular , plus and minus signs occurring rather promiscuously ; and while the corresponding seasonal means from the two periods agree in sign , the winter and summer means from the second period are numerically much larger than those from the first .
The lesser consistency of the V inequality is hardly surprising since it suffers from every uncertainty or accident that affects either the H or the I inequality .
A change of 10y in H alters I as much as a change of 24y in V. Thus the deduction of changes in V from the combination of observed changes in I and H , though the only way feasible , does not promise high accuracy .
There is a regularity in both sets of figures for H which can hardly be accidental .
They agree in making the successive monthly values all positive from May to August , and all negative from September to February .
Both periods show a clear maximum in the summer months , but one puts the minimum in winter , the other in equinox .
The inequality obtained by combining the two periods , allowing them equal weights , makes the winter and equinoctial values practically equal .
It shows a range of 12'5y , the maximum coming in June , and the minimum in November .
Table XVIII.\#151 ; Annual Inequality .
Month Horizontal force .
Inclination .
Vertical force .
and season .
First Second First Second A j / \ A M First Second AT period .
period .
iviean .
period .
period .
Mean .
period .
period .
y y .
y / r * y y y January . . .
-2-0 -1-8 -1-9 4-0-19 o-oo 4-0-09 4-1-9 - 4-4 -1-2 February . . .
-o-i -1-6 -0-9 + 0-03 4-0-06 4-0-04 4-0-8 - 1-8 -0-5 March -4-5 + 3-5 -0-5 4-0-41 -0-07 4-0-17 + 4-1 4- 6-0 4-5-0 April -2-3 + 0-5 -0-9 4-0-08 4-0-09 4-0-08 -2-6 4- 4-4 4-0-9 May + 3-1 + 5-6 + 4-3 -0-08 -0-17 -0-12 4-4-7 4- 7-3 4-6-0 J un + 8-8 + 7-0 + 7-9 -0-56 -0-16 -0-36 4-1-0 4-10-9 4-5-9 July + 2-6 + 2-3 + 2-5 -0-26 -0-18 -0-22 -3-0 - 0-8 - 1-9 August ... .
+ 4-4 + 0-5 + 2-5 -0-23 -0-07 -0-15 4-2-2 - 1-3 4-0-4 September . .
-4-2 -2-7 -3-4 4-0-12 -0-06 4-0-03 -5-6 - 8-5 -7-1 October ... .
-3-4 -3-7 -3-6 4-0-25 4-0-14 4-0-19 4-0-8 - 4-0 -1-6 November . . .
\#151 ; 1 * 9 -7-3 -4-6 4-0-21 4-0-36 4-0-29 4-3-0 - 4-4 -0-7 December . . .
-0-9 -1-7 -1-3 -0-16 4-0-06 -0-05 -8-0 - 2-2 -5-1 Winter ... .
-1*2 -3-1 -2-2 4-0-07 + 0*12 4-0-09 -0-6 - 3*2 -1-9 Equinox . . . .
-3*6 -0-6 -2-1 4-0-22 + 0*03 4-0-12 -0-8 - 0*5 -0-7 Summer . . . .
+ 4*7 + 3-8 4-4-3 -0-28 -0*14 -0-21 4-1-2 + 4-0 4-2-6 In the case of I the two periods agree in showing a distinct minimum in summer , but the one places the maximum in winter , the other in equinox .
Combining the two periods , allowing them equal weight , we get a comparatively smooth inequality with a range of ( f'65 , the maximum coming in November , the minimum in June .
The type is on the whole fairly similar to that of results that have been published for Parc St. Maur and Potsdam , and for the northern hemisphere as a whole by Liznar and Hann , but the range shown in Table XVIII .
is a good deal less than in the cases quoted .
Assuming that the phenomena are not of instrumental origin , it seems reasonably certain , after allowing for a secular change progressing at a uniform rate throughout the year , that H is higher and I lower in summer than in winter .
Also it seems probable , though more open to doubt , that V displays the same phenomena as H. This agrees with the results obtained from quiet days alone in a previous paper .
FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
S 17 .
The possibility that the phenomena owe something to instrumental causes must be recognised .
If there is any instrumental cause at work , it is presumably of thermal origin .
In the case of I , any sensible temperature effect on the absolute observations is difficult to imagine .
The strength of the dip needles , it is true , may not be the same , when they are stroked at 25 ' C. as when they are stroked at 10 ' C. I have no statistics on the subject .
Also the distance of the centre of gravity from the axis of suspension will naturally increase with rise of temperature .
Thus there would be nothing very surprising if the inclination observed with one particular end of the needle dipping varied slightly with temperature .
But the reversal of the poles , invariably observed at Kew , ought to eliminate this as a first approximation , leaving only second order terms , which one would not expect to be sensible .
The absolute observation hut at Kew is heated by a lamp in cold weather , so that the difference between summer and winter temperatures of observation is only of the order of 10 ' C. For the diurnal inequality in I we are dependent on the H and V magnetographs , the latter of which has a large temperature coefficient .
An error in the value accepted for that coefficient , or in the results accepted for the diurnal variation of temperature of the V magnet , might introduce a differential error as between summer and winter .
Any such error would affect the accuracy of the corrections applied to the observed inclinations to reduce them to the mean value for the day .
The hour at which the observations is taken is , however , one at which the change is not rapid either in the inclination or in the temperature of the magnetograph room .
Thus any considerable error seems improbable .
S 18 .
In the case of H , the ways in which temperature might come in are more numerous .
There might be an error in the temperature coefficient of the collimator magnet , but the consequences of this would not be nearly so serious as might appear at first sight .
What really concerns us in this connection is not the mean temperature of the whole H observation , but only the difference between the temperatures in the vibration and deflection experiments , which is usually only 1 ' or 2 ' C. In the present case , moreover , the error would come in not on the average size of this difference , but only on its seasonal variation .
Temperature practically always rises during an H observation , and so is higher during the deflection experiment , which comes last , than during the vibration experiment .
If there were no artificial heating the difference would naturally be greatest in summer , so a seasonal differential error is conceivable .
If it existed , however , it should show itself in an apparent seasonal variation in the values found for the magnetic moment of the collimator magnet reduced to 0 ' C. , and a special investigation showed no trace of this .
Errors in the values assumed for the variation with temperature of the length of the deflection bar and the moment of inertia of the collimator magnet would come in on the full annual range of temperature in the magnetic hut , but coefficients of thermal expansion in brass and steel are small quantities , and any large percentage error in them is most improbable .
The temperature coefficient of the H magnetograph is small , and the general accordance of individual H observations and curve measurements seems to preclude the possibility of any large percentage error in the value accepted .
There is , in short , no direct evidence of the existence of uncorrected temperature effects , while there is a good deal in favour of the substantial accuracy of the allowances made .
At the same time , direct experiment , if possible without undue risk of interfering with the records , would be desirable .
Fourier Coefficients .
S 19 .
The analysis of the diurnal inequalities into series of harmonic terms whose periods are 24 , 12 , \amp ; c. , hours is part of the regular routine at some observatories .
To some people circular functions seem on a wholly superior plane to all others , and the analysis of diurnal inequalities according to any other type of function would appear almost inconceivable .
Others may argue that the employment of circular functions for the representation of magnetic diurnal inequalities may hide rather than reveal what is of real physical importance .
There are several common-sense arguments in favour of the analysis in series of sines and cosines .
The use of these is so very general that for the intercomparison of results at different stations there is really no competing series .
Again , it is found that as a rule 4 Fourier waves\#151 ; i.e. , terms with periods of 24 , 12 , 8 and 6 hours\#151 ; suffice to give a very close approximation to the observed diurnal inequality , and the 8-hour and 6-hour waves are usually small compared with the first two .
There is thus a good deal to be said for the argument that the Fourier analysis is a natural one .
This argument of course would lose much of its force if a single function of the time , of a not unduly complicated form , with only 2 or 3 parameters sufficed to represent the diurnal inequality adequately at a number of stations .
But until such a function has been proved to exist , the course which is followed here of employing ordinary Fourier series is likely to commend itself to the majority .
The diurnal inequality is usually expressed in one of the alternative forms ax cos t + bisin t + a2 cos 2 sin ... Ci sin ( t + a.i ) + c2 sin ( + + ... , where t represents time counted from midnight , one hour in t being taken as the equivalent of 15 ' .
The constants a , b of the first series , or c , a of the second , are generally called Fourier coefficients , c representing the amplitude and a the phase angle .
The usual process is to calculate the a , b constants from the hourly values in the diurnal inequality , and then deduce the c , a constants from the equations tan a = a/ b , c = \/ a2 + b = sin = cos If a , b coefficients have been calculated for three magnetic elements , they can be deduced at once for any other element , through the formula expressing it in terms of FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
the first three .
Thus it sufficed to calculate the coefficients in the usual way for D , H and V , and the work so far as D is concerned had been already done .
The values of a and b for the year , or for any season , are the arithmetic means of the values from the included months .
They can thus be deduced at pleasure from the values of a and b calculated for the included months , or be specially calculated from the inequality for the season .
The employment of both methods puts the accuracy of the calculations to a very severe test .
In fact , to secure agreement to say three significant figures it is necessary to carry the calculation at least one place further .
This test was not in general applied , but all the calculations were carefully checked , special attention being given to any apparent irregularities .
Table XIX.\#151 ; Horizontal Force ( Local Mean Time for Phase Angles ) .
Year .
Cl. OCi .
C2\#187 ; a2 .
a3* C4 .
a4 .
7-\#166 ; o 7- o 7- 0 7- 0 1890 6-81 119-2 4-01 312-4 2-18 175 .
1-13 28 1891 9-66 117-1 5-62 303-7 2-37 166 1-02 23 1892 13-00 117-2 6-52 301-6 2-81 166 1-32 22 1893 13-35 114-1 7-45 301-7 3-26 163 1-33 18 1894 13-26 113-9 7-09 294-9 2-60 167 1-26 24 1895 11-73 113-3 6-15 298-9 2-57 171 1-20 22 1896 9-81 114-8 5-22 297-2 2-48 166 1-24 17 1897 8-56 112-3 4-61 294-8 1-99 173 1-30 28 1898 8-36 115-9 4-14 302-1 2-17 166 1-04 25 1899 8-10 115-8 4-09 *- 302-5 2-36 174 1-10 32 1900 6-56 119-0 3-85 306-7 2-16 178 1-12 20 Tables XIX .
and XX .
give the amplitudes and phase angles for the 24- , 12- , 8- and 6-hour terms in the H and V mean diurnal inequalities for the whole year , each of the eleven years being treated separately .
Table XX.\#151 ; Vertical Force ( Local Mean Time for Phase Angles ) .
Year .
Cl. \#171 ; 1- a2 .
Cs .
0C3 .
Ci- a4 .
7- 0 7- 0 7 0 7 0 1890 5-08 139-4 3*39 273-3 1-36 106 0-47 282 1891 6-84 145-1 4*79 272-1 1-64 96 0-60 295 1892 8-10 148-8 5 .
29 268-5 1-77 101 0-69 280 1893 7-26 138-8 5*62 270-1 2-26 97 0-71 279 1894 8-38 145-8 5*32 269-0 2-02 95 0-60 292 1895 7-70 148-1 5*09 271-6 1-83 97 0-68 295 1896 7-54 143-5 4*60 270-6 1-78 105 0-61 284 1897 7-07 138-5 3*95 268-6 1-48 101 0-52 286 1898 6-95 142-7 4*25 274-2 1-50 108 0-49 286 1899 6-00 144-0 3*87 271-3 1 - 31 107 0-52 284 1900 5-38 130-0 3*48 276-1 1-60 103 0-61 281 G.M.T. was employed , as already stated , in the diurnal inequalities , and thus the values deduced in the first instance for the Fourier coefficients referred to G.M.T. Local mean time is , however , more generally instructive , and accordingly the necessary transference to local time has been made in Tables XIX .
and XX .
, and all the subsequent tables relating to Fourier coefficients .
The transference does not affect the amplitudes .
The corrections it requires to the phase angles are - .
+ 19 ' in a1 } +38 ' in a2 , + 57 ' in as and +1 ' 16 ' in a4 , Kew local time being the equivalent of 19 ' after Greenwich .
The relation with sunspots will be discussed more fully later , but the significance of the tables will be better understood if regard is paid to the following facts .
1890 was a year of sunspot minimum with a frequency of 7T .
The frequency rose rapidly to a maximum of 84'9 in 1893 , and fell more slowly to 9'5 in 1900 .
In 1898 there was a slight arrest in the fall , the mean frequency being 26'7 as compared with 26'2 in 1897 .
The year 1893 , though the year of sunspot maximum , was on the whole a quiet year free from large disturbances , much quieter than 1892 , 1894 or 1895 .
In Table XIX .
cu c2 , c3 and c4 all rise to a maximum in 1893 , and the general , parallelism with the sunspot variation is readily seen , especially in c1 and c2 .
Table XX .
shows a similar state of matters , except that the value of c4 in 1893 fell short of that in several adjacent years .
The variation in the phase angle is by no means so decisive .
In Table XIX .
the largest values of a4 and a2 are associated with the years of sunspot minimum , but the smallest values occur in 1897 , an intermediate year as regards sunspots .
Accident seems to play a sensible part in the relation between phase angles in successive years , especially in the case of a3 and a , .
The same remark applies in even greater measure to Table XX .
In fact the influence of sunspot frequency on the phase is so small that for its study it is desirable to combine the years in groups representative respectively of large and small sunspot frequency , in hopes of eliminating accidental features .
S 20 .
Tables XXI .
and XXII .
show the variation in amplitude and phase angle throughout the year in the Fourier waves representing the diurnal inequality in H , and Tables XXIII .
and XXIV .
do the same for V. Results are given for the whole 11 years , also for 1890 , 1899 and 1900 representing few sunspots , and 1892 to 1895 representing many sunspots .
Arithmetic means from the 12 months are given in Tables XXI .
and XXIII .
When there is considerable variation of phase throughout the year , contributions from different months to the seasonal and yearly diurnal inequalities to some extent neutralise one another .
Thus frequently a better idea of the average activity of the forces to which any particular Fourier wave is due is derivable from the arithmetic mean of the c 's than from the corresponding c in the mean diurnal inequality for the year .
FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
Table XXI.\#151 ; Horizontal Force .
( Unit lie .
) *1 .
^2* cs .
C4 .
Month .
S. mini- mum .
11 years .
S* .
maxi- mum .
s. mini- mum .
li years .
S* .
maxi- mum .
s. ; mini- mum .
11 years .
maxi- mum .
s. mini- mum .
11 years .
s. maxi- mum .
January . . .
1-23 3-15 5-41 1-80 3-41 4-46 2-27 2-26 2-52 1-16 1-29 1-33 February . . .
2-93 4-50 6-33 2-46 3-54 5-02 2-45 2-68 3-14 1-31 1-26 1-18 March ... .
6-55 9-13 11-78 4-31 5-66 7-46 3-39 3-75 4-53 1-44 1-63 1-88 April 10-27 14-43 18-44 5-37 7-55 9-20 3-17 3-61 4-17 1-32 1-65 2-09 May 12-21 15-83 19-14 4-94 6-44 8-11 1-32 1-25 1-57 0-86 1-04 1-18 June 13-18 17-18 21-56 5-58 7-06 9-26 1-85 1-73 2-45 0-88 0-71 0-96 July 12-94 17-17 21-86 4-83 7-06 9-10 2-36 2-01 1-99 0-95 0-73 0-88 August . . . .
12-41 15-97 19-70 4-74 6-35 7-90 3-12 3-18 3-44 1-40 1-43 1-32 September . .
10-60 13-17 15-31 5-05 5-71 6-33 3-63 3-84 3-96 1-84 1-88 1-97 October . . .
8-03 10-89 13-88 4-88 6-14 8-02 2-72 3-66 4-30 1-13 1-62 1-91 November . .
3-05 5-20 7-98 3-72 4-54 5-68 1-85 2-21 2-61 0-92 1-13 1-18 December . .
2-66 3-30 4-73 2-59 3-32 4-25 1-53 1-62 2-07 0-67 0-57 0-46 Arithmetic 1 means .
J 8-01 10-83 13-84 4-19 5-56 7-07 2-47 2-65 3-06 1 -16 1-25 1-36 In the case of cx and c2 , whether in Table XXI .
or Table XXIII .
, there is no single month in which the amplitude fails to show an increase as we pass from the sunspot minimum to the sunspot maximum group of years .
In the case of c3 and c4 the same is generally true , but there are a few exceptions , presumably accidental in both tables .
Table XXII.\#151 ; Horizontal Force .
Phase Angles referred to Local Mean Time .
Month .
\#171 ; ! .
a2 .
\lt ; *3 CI4 .
S. mini- mum .
11 years .
S. maxi- mum .
\ S. mini- mum .
11 years .
S. maxi- mum .
s. mini- mum .
11 years .
S. maxi- mum .
S. mini- mum .
11 years .
S. maxi- mum .
0 / 0 / 0 / 0 / 0 / 0 / 0 0 0 0 0 0 January -24 24 51 43 67 12 298 41 282 8 279 30 167 5 157 *9 152 -1 8*5 3 5 -2*7 February 69 21 70 49 77 46 280 32 269 56 271 57 146 -7 146 3 141 -2 -9*4 -5*2 5 '9 March 103 28 101 8 102 29 297 55 294 15 295 28 154 -0 153 5 153 '6 0*4 -1 *4 -6*2 April 115 32 114 51 115 30 299 0 297 51 297 45 145 3 150 -1 152 *3 13 -1 20 -4 20 -3 May 132 15 131 27 130 17 317 28 307 52 302 44 217 *0 204 *5 198 *5 59 -9 67 *0 63 -3 June 131 37 134 43 136 25 306 54 308 24 312 5 200 *8 213 3 216 0 37 -0 53 -3 67 *2 July 134 16 134 5 134 20 321 2 310 37 307 59 204 -5 191 9 179 9 42 1 30-1 30 *8 August 129 21 127 58 126 44 338 46 327 27 322 17 212 *4 203 *8 197 -8 32 -2 33 *9 24 *7 September ... .
116 18 113 58 111 58 320 32 320 48 318 10 187 -8 188 *3 188 -7 34 6 33 5 38 0 October 94 47 94 30 93 43 294 57 295 57 297 8 162 -8 162 *9 160 -8 25 *4 21 *3 16 *2 November ... .
80 37 76 20 77 37 289 21 282 2S 279 45 158 9 153 9 149 -1 40*7 24 -0 6*8 December ... .
13 50 37 45 54 31 280 16 268 59 265 50 145 6 141 -8 134 -2 34 *3 15 *8 4-4 The annual variation in the amplitude is clearly shown in all cases , but it follows different laws in the different waves .
Ci and c2 in both H and V have a well marked vol. ccxvi.\#151 ; a. 2 L minimum near midwinter and a maximum near midsummer .
c3 and also exhibit a minimum near midwinter , but there is a second minimum\#151 ; which in the case of c3 in Table XXIII.\#151 ; Vertical Force .
( Unit lie .
) Cl. ^2\#187 ; Cs- U4 .
Month .
S. mini- mum .
11 years .
s. maxi- mum .
S , mini- mum .
li years .
s. maxi- mum .
s. mini- mum .
11 years .
s. maxi- mum .
s , mini- mum .
11 years .
S. maxi- mum .
January . . .
2-72 3-53 4*39 1-10 1-32 1-70 0-51 0-57 0-71 0-28 0-44 0-54 February . . .
3-88 5-01 5-98 1-94 2-72 3-29 1-03 1-31 1-58 0-36 0-47 0-63 March ... .
5-70 7-03 7*26 4-21 5-37 6-24 2-17 2-44 2-78 0-94 0-98 1-04 April 7-81 8-39 8*50 5-38 6-49 7-47 2-24 2-75 3-03 0-70 0-82 0-87 May 9-28 10-88 12-18 6-02 7-62 9-26 1-93 2-10 2*25 0-52 0-63 0-60 June 8-49 10-33 11-49 5-28 6-65 8-22 1-27 1-42 1-80 0-47 0-36 0-32 July 8-76 1107 13-54 5-12 6-38 7-55 1-40 1-66 2-01 0-43 0-37 0-37 August . . . .
5-61 7.37 7-78 5-27 6-21 6-75 2-05 2-47 2-89 0-62 0-62 0-64 September . .
5-89 7-64 8-35 3-90 4-86 5-34 1-96 2-07 2-25 0-85 0-81 0-91 October . . .
3-90 5-78 7*27 2-07 3-49 4-17 1-59 1-89 2*29 O'86 1-04 1-18 November . .
308 4-47 5-32 1-61 211 2-64 0-74 1-00 1-32 0-61 0-68 0-80 December . .
1-79 2-95 4-02 0-88 1-26 1-78 0-53 0-78 1-06 o-io 0-21 0-41 Arithmetic 1 means .
j 5-58 7-04 8-01 3-56 4-54 5-37 1-45 1-70 2-00 0-56 0-62 0-69 Table XXI .
is actually the principal minimum\#151 ; near midsummer .
The largest values in c3 and c4 present themselves in equinoctial months .
Table XXIV.\#151 ; Vertical Force .
Phase Angles referred to Local Mean Time .
Month .
*i- a2 .
a3 .
a4 .
S , mini- mum .
11 years .
S. maxi- mum .
S. mini- mum .
11 years .
s. maxi- mum .
mini- mum .
11 years .
S. maxi- mum .
S. mini- mum .
11 years .
s. maxi- mum .
January . . .
0 164*1 0 160-8 0 157-9 0 287-9 0 283-4 0 275-0 0 145 O 118 0 101 0 267 0 277 0 281 February . .
154*6 155*7 154-5 269-8 275-9 278-1 104 100 98 237 261 282 March ... .
133-3 143-9 150-3 272-7 270-7 268-6 94 93 87 268 275 282 April 121*8 130-5 132-1 265-4 263-3 260-3 95 94 92 293 300 298 May 126*9 129-3 129-5 274-9 270-3 268-4 103 102 99 302 314 327 June 136-2 139-5 140-9 271-2 268-9 268-3 107 100 98 307 315 307 July 145*2 142*8 146-3 266-3 263-6 262-7 91 82 78 260 245 219 August . . . .
128-8 134*5 136-3 276-8 273-5 273-5 104 102 101 285 280 281 September . .
139*4 143*1 145-3 279-4 276-6 277-7 116 109 108 290 297 298 October . . .
141*7 149*5 153-6 267-2 269-5 267-6 102 102 95 281 287 284 November . .
170*4 171*8 174-6 308-7 297-1 292-5 138 125 108 302 286 278 December . .
143-0 155*9 163-7 281-2 274-3 269-4 125 119 120 306 267 269 FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
In the case of the phase angles the difference between sunspot maximum and minimum years is much less decisive , a3 in V in every month of the year is larger\#151 ; i.e. , speaking generally , the maximum occurs earlier in the day\#151 ; in sunspot minimum than in sunspot maximum .
There is also a large preponderance of months in which the sunspot minimum angle is the larger in the cases of a3 in H and of a2 in both H and V. aj , however , in V shows exactly the opposite phenomenon , while a , in H and a4 in both H and V show no decided tendency .
In H all four phase angles exhibit the same tendency in their annual variation .
They are distinctly larger in summer than in winter .
This is especially conspicuous in the case of a4 ; in sunspot minimum the phases of the 24-hour term in January and July approach opposition .
An increase in phase angle means in a general sense an earlier occurrence in the maximum and minimum , but this requires special interpretation at times .
Take , for instance , the 11-year values of oq in H. In January and February the values are respectively 51 ' 43 ' and 70 ' 49 ' .
The corresponding times of occurrence of the maximum are respectively In January t = ( 90 \#151 ; 51'72)/ l5 = 28'28/ 15 = T89 hour = 1 hour 53 minutes .
In February t = ( 90 \#151 ; 70'82)/ l5 = 19T8/ 15 = 1'28 hour = 1 hour 17 minutes .
But when we pass to March the phase angle 101 ' 8 ' falls in a different quadrant , and the time of the maximum is given by t = ( 450\#151 ; 101T3)/ l5 = 23-3 hour = 23 hours 18 minutes .
In a sense the maximum has become earlier in March , only it has as it were transferred itself to the previous day ; the wave , in fact had already passed its maximum when the day commenced .
The minima in the three months occur in January at 13 hours 53 minutes , in February at 13 hours 17 minutes , and in March at 11 hours 18 minutes .
Thus the statement that the minimum has become earlier as the phase angle increased was in this case literally true .
In the case of V in Table XXIV .
the annual change is on the whole in the opposite direction to that in H , the angles a1 } a2and a3 being all smaller in summer than in winter .
There seems to be a very appreciable accidental element in the values obtained for individual months , especially in the case of a , .
The mode of annual variation of the amplitudes of the several Fourier waves is best shown by expressing the values for different months as fractions of their arithmetic mean .
This has been done for the full 11-year results in Table XXV .
, data being given for both II and V. The laws of annual variation for the two elements proved sufficiently alike to encourage the formation of the arithmetic means from the two .
These arithmetic means are considerably smoother than the data from H or V alone .
Relatively considered , the amplitudes of the first four Fourier waves show a fairly similar range of annual variation .
On the whole the range is largest in c , and least in c3 ; but c3 and c4 , having a double annual period , vary most rapidly .
Table XXV.\#151 ; Ratios to Arithmetic Mean ( 11 years ' data ) .
Month .
Cl. .
c^\#187 ; H. V. Mean .
H. y. Mean .
H. Y. Mean .
H. V. Mean .
January . . .
W 0-29 0-50 0-40 0-61 0-29 0-45 0-85 0-33 0-59 1-03 0-72 0-87 February . .
0-42 0-71 0-56 0-64 0-60 0-62 1-01 0-77 0-89 D01 0-75 0-88 March ... .
0-84 1-00 0-92 1-02 1-18 1-10 1-41 1-43 1-42 1-31 1-58 1-45 April .
... .
1-33 1-19 1-26 1-36 1-43 1-39 1-36 1-61 1-49 1-32 1-32 1-32 May 1-46 1-55 1-50 1-16 1-68 1-42 0-47 1-23 0-85 0-83 1-02 0-93 June 1-59 1-47 1-53 1-27 1-46 1-37 0-65 0-83 0-74 0-57 0-58 0-57 July 1-59 1-57 1-58 1-27 1-41 1-34 0-76 0-97 0-87 0-59 0-60 0-59 August . . .
1-48 1'05 1-26 1-14 1-37 1-25 1-20 1-45 1-32 1 .
15 1-00 D08 September . .
1-22 1'09 1-15 1-03 1-07 1-05 1-45 1-21 1-33 1-51 1-31 1-41 October . . .
1-01 0-82 0-91 1-10 0-77 0-94 1-38 Ml 1-25 1-30 1-68 1-49 November . .
0-48 0-64 0-56 0-82 0-46 0-64 0-84 0-58 0-71 0-91 1-10 1-01 December . .
0-30 0-42 0-36 0-60 0-28 0-44 0-61 0-46 0-53 0-46 0-34 0-40 S 21 .
Fourier coefficients were not calculated for the individual months of the year except for H and Y. For the other elements they were calculated only for the diurnal inequalities from the seasons and the year .
Table XXVI .
compares the amplitudes in these seasonal diurnal variations for H , V , T , N , W , and I , use being made in dealing with N and W of the results previously obtained for D. In the case of cx and c2 the equinoctial value is always intermediate in size between the winter and summer values , and somewhat in excess of the value for the whole year .
In the case of c3 and c4 the equinoctial value is invariably in excess of both the winter and summer values , and much in excess of the value for the year .
In the case of the summer* value is the lowest for two elements , N and H ; in the case of c4 the winter value is less than the summer value in no element except I. In the case of all the Fourier waves for the three rectangular components V , N , and W , the year value of the amplitude is largest in W and least in V. This is in general true also of the three seasonal values , but in equinox c4 is larger in N than in W , and there are one or two cases in which the Y value is not the lowest .
Table XXVII .
shows the ratios borne by the amplitudes of the 12- , 8- , and 6-hour waves to the amplitude of the corresponding 24-hour wave .
In the case of H , T , N , and I , the importance of the 12- , 8- , and 6-hour waves falls relative to that of the 24-hour wave as we pass from winter to equinox , and from equinox to summer ; but in W and V the relative importance of the shorter period waves is greatest in equinox , and in the case of the 12-hour wave it is least in winter .
Table XXVI.\#151 ; Amplitudes of Fourier Waves ( 11 years ' data ) .
( Unit lie in Force Components .
) FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
uouirang N io 04 co -h Oi ^ ^1\gt ; 00 O o o o o o ^ 4 .xoumbg ; CO OO^OOOJrH CO G5 O H H r\#151 ; t O r-H r\#151 ; 1 04 ^ U9 ?
UtAV rH Tf^ CO N o N O IQ 1\gt ; CD O r\#151 ; 1 O O O rH ^ 04 00 05 i\gt ; - 04 00 h lO CO 00 CD O rH O O O rH ^ J9iuumg 00 CO 05 H Oi CO lO O 00 U- rH 00 r\#151 ; H 04 rH rH J-H ^ \#163 ; .xouinbjj o 05NIOHOC4 50 04 05 04 00 04 CO 04 04 CO lO q .J9BIIA\ .
CD 00 O 00 50 CD 04 h D iO 04 CO h 04 O rH 04 04 ^ CO 00 00 00 O IQ T^CDOHCOH 04 m 04 04 ^ .joixirang CD O 04 05 CO .
CD 04 05 CO CD CD 00 N H i- .
rH w .xoumbg ; 04 04 CO CD CD rH O O O rH rH CO CD LO CD 1\gt ; O q .jo^uiav iO 00 CO D D o D CD 00 O 04 CD rH CO H CO ^ ^ 50 H Cl H CD lO C0 10OC0N04 iO r-H CD CD 00 \#163 ; 0 uouiuing 05 04 N ^ H ( M \lt ; M 0 00^10 10 00 D D lO CD DX rH rH i-H i\#151 ; 1 W rH .xoumbg ; 05 N CDN CD lO 04 N H D D D CD H N O ^ CO X rH rH rH rH W Vi 00 O CD 04 CD 50 05 05 05 N N CD 04 CO CO CO CD 00 0 o ^ r\#151 ; 1 04 rH CO O 05 05 05 rH rfl iO 05 tD 05 04 CO Xs rH i\#151 ; 1 W \#169 ; a CD M K* H ^ l-H r-H Summer .
CD 4 CO H 4 D o o o o o o \#169 ; \#169 ; \#169 ; \#169 ; \#169 ; o o H CO O 00 00 00 .5 rH rH rH ( ^\gt ; rH rH *3 / *-H \#169 ; \#169 ; o o \#169 ; \#169 ; w X \#169 ; h h lO h 0O 4 04 rH rH rH rH 04 Jh o \#169 ; o \#169 ; \#169 ; \#169 ; gS \#163 ; .4-5 c3 ."0 04 05 04 CD ~aQ c3 rH \lt ; 0 ^0 rH rH X o3 05 \#163 ; o \#169 ; o o o \#169 ; \lt ; X\gt ; r-H X rH CD 04 05 H t\gt ; - 50 05 v^ a rH rH rH 04 rH go \lt ; v o o o \#169 ; o \#169 ; \gt ; m \#163 ; h o O 04 00 04 04 lO .2 * JL| 5 '3 qh CO CO 04 ( M H CO \#169 ; \#169 ; \#169 ; \#169 ; \#169 ; \#169 ; 2 o A w .
^4H X 05 \#169 ; CO CO CO N 04 o 50 04 CO 04 ^ GO .3 \#169 ; \#169 ; o o \#169 ; o 0 T3 S3 4-5 r-H X o3 50 H oo 04 O Ph 04 04 04 rH CO CO a \#169 ; !
* \#169 ; o \#169 ; o o \#169 ; o X gq \#169 ; O 00 CO Jt\gt ; \#187 ; rH rH O a H \#169 ; iQ 4 \#169 ; H *\#166 ; S a o o \#169 ; o o o m 1 .
hH M o 04 O CO 05 CO 00 hH C3 iOi\gt ; \#169 ; 4N 4 t\gt ; '3 o \#169 ; \#169 ; \#169 ; o \#169 ; M X H H c w " *^T X \#169 ; 4 \#169 ; CO CO CO 50 05 4 00 \#169 ; 50 \#169 ; \#171 ; g 6 6 \#169 ; o \#169 ; o H \#163 ; X c3 tH 50 rH 04 50 rH 50 CD CD 50 CD 50 \#169 ; O O O O O O H P fl u-\gt ; a CD W t\gt ; H ^ ^ H-1 i Table XXVIII.\#151 ; Amplitudes of Fourier Waves .
( Unit lie .
) DISCUSSION OF KEW MAGNETIC DATA \i9inumg COHOOQ O O CO HH r-H r-H O O .xoumbg ; a\gt ; 05 05 ( M GO CO 05 00 r-H i\#151 ; 1 O O HH lO 05 O o 05 io co rb \#169 ; O O .a'eoj^ Ib O to HH CM )\#151 ; i CD IO r-H r-H \#169 ; \gt ; \lt ; \#169 ; .jouimng r-H lO r-H lO CO H N C\#163 ; ) ( M ( M ( NH .xoumbg ; rH CO ib CD r-H O IO 05 HH CO Cl H N O IO N lO O rH CD CM CM r-H o H ^NH 00 CM 05 HH \lt ; M CM r-H r-H .jouming CO CO CO r-H IQ 05 05 HH 00 HH Ib IO $ .xoumbg ; CD Cl N N o ooi\gt ; oo Ib HH IO CO .J^uiAV co co co co 00 CD CO CO HH CM CM r-H O CM CO CM 00 O CO io \lt ; D HH IO CO .joraumg 1\gt ; 00 GO b-lO CD H 05 OClHb ( Mi\#151 ; 1 r-H .xoumbg D lO CD N CO b Nb HH 00 Ib IO r-H HH lb 00 co o 05 oo oo CO h Hf Cl .at ?
aA CO r-H HH HH 00 CM 00 HH CM b b IO r-H Period .
Sunspot maximum " minimum " maximum " minimum .guQiuQjg ; EEbb GQ \lt ; D \gt ; Sh .2 ' \#163 ; 2 o bH o m \lt ; r\gt ; \lt ; 4h O cn O \#171 ; pH M hH X X w i-i PQ \lt ; 1 H .J9uuung IO oo CO CD o o o o o o o o A .xoumbg CO CD CO HH r-H r-H i-H r-H o o o o .aa^uijw lb 00 CM r-H rH HH rH rH o o \#169 ; o U'B 9^ OlOOOO r-H r-H O rH \#169 ; o o o .J9uimng r-H b- O r-H r-H rH CM CM O O O O rH .xoumbg 00 io CO HH CM CO CO CO O \#169 ; O O ua^uiAV CO H HH HH HH o CM CM O r-H O O CM rH IO CD CM CO CM CM o o o o .J9uiiung rH 05 rH 00 H CO b CD o \#169 ; o o A .xoumbg CM IO HH ib lO lO b CO \#169 ; \#169 ; \#169 ; \#169 ; uaUiiM O CO 00 ib 00 co HH HH \#169 ; ^H o o *ai ?
9^ CO CO H lO io IO IO CD \#169 ; \#169 ; \#169 ; \#169 ; Period .
i \#151 ; Sunspot maximum . . .
" minimum . . .
" maximum . . .
, , minimum . . .
.(jueraoig EEbb FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
Table XXVIII .
contrasts the sunspot maximum and minimum values of the amplitude for the year and seasons in both H and V. In H the sunspot maximum value of the amplitude is always the greater .
Its excess , however , relatively considered , diminishes as the order of the harmonic increases , and except in it is least marked in summer .
In V the sunspot maximum value is also always the greater , except in the case of the summer value of c4 ; but the pre-eminence of the sunspot maximum value is not specially conspicuous in c4 .
Table XXIX .
contrasts the ratios borne to c4 by c2 , c3 , and c4 in the seasonal diurnal inequalities in the sunspot maximum and sunspot minimum groups of years .
In the case of V the ratios seem little dependent on sunspot frequency , whereas in H with one exception they rise as sunspots diminish , especially in winter .
In that season , in fact , in sunspot minimum the amplitude of the 24-hour term is exceeded by that of the 12-hour term and even by that of the 8-hour term .
S 22 .
Table XXX .
contrasts the phase angles in the diurnal inequalities for the year and seasons in the six elements H , V , T , N , W , and I. An increase in the phase angle as we pass from winter to equinox , and from equinox to summer , is shown by H , N , and T in all four Fourier waves .
It is also shown by V in the case of a4 , and by W in the case of a2 .
A fall as we pass from winter to equinox , and from equinox to summer , is shown by V in the case of alt and a3 , and by W in the case of a4 .
T is remarkable for the smallness of the seasonal variation in the phase angles , especially in the case of a2 .
Table XXXI .
contrasts the seasonal and yearly diurnal inequality phase angles in years of sunspot maximum and minimum .
In the case of a2 the sunspot minimum angle is invariably the greater for both H and V. The same is true of a3 , except in the case of H in equinox , when the sunspot minimum angle is slightly the smaller .
The same rule is observed* by a4 in H , except in summer ; but in V the sunspot maximum value of a4 is the larger , though its excess is generally small .
The sunspot maximum value of a4 is the larger in summer and winter in H , and in all cases in V. Table XXXII .
shows the difference between the phase angles found in the present paper for ordinary days and those found in a previous paper* for quiet days .
It gives the excess of the ordinary day over the quiet day phase angle converted into time at the rate of 1 hour = 15 ' in ah 30 ' in a2 , 45 ' in a3 , and 60 ' in a4 .
The plus sign is equivalent to an earlier occurrence of phenomena in ordinary days .
The values under the seasons , it should be noticed , are from the seasonal inequalities , and are not arithmetic means from the months included .
The outstanding feature is the preponderance of plus signs in the case of the 24-hour term in V , and the large size of most of the differences from this term .
* ' Phil. Trans. , ' A , vol. 202 , p. 335 .
Table XXX.\#151 ; Phase Angles Referred to Local Mean Time ( ll years ' data ) .
DR. C. CHREE : DISCUSSION OF KEW MAGNETIC DATA .jouiumg 0 lOHCCWCCh H Oi H H lO ( N cq co cq .xoumbg ; HOilOOOHh 0 Oi OO CO ( M CO H rH 00 Cl | 50 rH cq co 1 cq .J91U ?
AY 00050000 0 OOlOOiOhh x\gt ; - cq i\#151 ; -i 'ht 05 oq CO | rH *JR9^ co O co CO H H 0 cq CD GO 50 H H cq oo cq wh cq co cq eo .J9inumg OCOCOiO(Mh 0 COh co 05 CD H O 05 Cl 1C CO H cq rH rH cq .xoumbg oo co x^ cq ^ rH 0 co oo co lO ( M H cd 05 cq co cq , H rH i\#151 ; 1 cq , J9IUIAV HNHO5CO00 0 o co cq h oo ci io h co co cq cq rH i-H rH rH Cq | .JR9^ 05 OO N H CD O 0 05 O 50 00 00 CD co o cq co cq rH rH i\#151 ; t r\#151 ; i Cq a .J9uiumg " O lO 00 \lt ; M H t^ cq CO 50 rH 50 CO 05 H H H io 0 H CD 00 oo H IO co cq cq cq i-h .xoumbg ^ H cq co h cq o tH co h 'rh cq H 05 O \gt ; C H CD 0 o CD OO N CO CO co cq cq cq rH .J9BitAY " ( NCDCOlOlOO Gq no 50 h H CD CM 05 1\gt ; H H o l\gt ; 00 1\gt ; lO H 05 cq cq cq cq *JR9^ v CO CD O 00 CO 00 50 50 rH 50 50 OhOOCIH 0 cu\gt ; OOh rt CO co cq cq cq rH A .J9uiumg v 05 0 00 00 05 0 H H co cq H cq CD H H H o O CO CO CO I-H O rH rH rH1 1\#151 ; H rH CM CO .xoumbg ^ O-CD CO CO CO h cq o cq h H h NOhHHh o o h cq 05 cq i\gt ; H rH rH ( M Cq , J94UJAY ^ CO rH 00 rH CO cq rH Cq 50 rH 50 ClHhl0 05 0 o co co co co -hi cq rH rH cq cq o / 115 23 142 43 132 33 95 48 219 11 285 15 Element .
S !
\gt ; H !
z ; pS~ 0 a \amp ; cS 0 a IS O o O '"d 0 P u eg 0 Ph go 0 q 0 GO hH H H M H a PQ ?
Eh \#163 ; .J9uirang 00 CO cq rH O 50 rH O 05 h h 05 OO cq cq .xoumbg co 50 cq o h 05 05 cq rH rH 00 00 cq cq .j^uiay cq co co cq o CO hH oo oo H h h cq cq .JB9A 05 H* co co o o ''h CO CO cq cq oo oo cq cm .J9TUUing 05 50 05 50 o 00 00 ^ H 05 O 05 O rH cq rH .xoumbg cq oo cq o co cq h h CO CO 05 O rH i-H rH 'J91UFAY 50 H CO o HH 50 CO Hh h 50 O Cl rH rH i-H rH .JR9^ X\gt ; - 05 H 05 o CO CO tH COi\gt ; 05 O H rH rH \#163 ; uoraumg " h lO h CO cq cq rH o oo cq o Hcqcoh co co cq cq .xoumbg " OiOOOh H rH 'hH rH CO t\gt ; rH o OOCOh co co cq cq .jg^uiAY " 00 O CO o CO hH 50 co CO 05 CO o i\gt ; ooi\gt ; oo cq cq cq cq .JR9^ O / 299 15 306 36 269 43 273 42 rH .joraiung o / 132 15 131 53 138 39 134 39 .xoumbg o / 106 57 108 57 144 50 132 24 .jg^uiAY ^ 50 05 CO 00 50 rH CO rH o oo cq 05 O h h O 50 rH i-H .JRO^ o / 114 40 117 31 145 31 137 59 Sunspot .
Maximum . .
Minimum . .
Maximum . .
Minimum . .
.{JU9UI9ig KK\gt ; \gt ; FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
There is obviously a considerable " accidental " element in the results for individual months , especially in the shorter period-waves .
There is , however , an unmistakable seasonal variation in the phenomena in the 24-hour term in H. In the summer months , the difference , though small , is clearly positive , whereas in winter it is not merely negative but very large , exceeding an hour of time .
The fact is the seasonal variation in ai in H is much more pronounced in ordinary than in quiet days , the monthly values showing a range of 97 ' in the first case as compared with 71 ' in the second . .
Table XXXII.\#151 ; Ordinary less Quiet Day Phase Angle in Minutes of Time .
Month and season .
24-hour term .
12-hour term .
8-hour term .
6-hour term .
H. V. H. V. H. V. H. V. January . . . .
68 - 5 + 8 -38 - 4 + 5 - 5 - 8 February . . .
- 60 + 16 - 1 -23 - 4 + 6 + 6 + 3 March - 2 + 95 + 1 - 3 - 4 + 2 - 2 - 4 April + 19 + 76 + 7 - 3 + 3 + 1 + 13 + 7 May + 8 + 66 -14 - 9 + 5 - 1 + 4 + 14 June + 3 + 48 - 9 - 3 + 8 -13 + 24 + 7 July + 12 + 58 - 2 - 5 + 15 - 1 + 18 + 10 August ... .
+ 5 + 60 + 1 - 9 - 1 - 3 - 3 - 9 September . . .
31 + 100 -25 - 9 -14 - 2 + 1 -14 October ... .
\#151 ; 25 + 112 + 1 - 7 0 - 1 + 4 + 8 N ovember . . .
-- 69 + 55 - 7 -26 - 4 + 14 0 - 3 December . . .
\#151 ; 100 + 67 - 1 -64 - 5 + 6- + 14 -17 Year 6-2 + 72-2 - 3-0 - 9-7 - 5-3 + 0-2 + 3-8 + 0-1 Winter ... .
- 82-6 + 29-8 - 1-8 -15-8 - 4-5 + 8-1 + 2-2 - 4-8 Equinox . . . .
\#151 ; 9-0 + 94-9 - 3-4 - 5-5 - 4-7 0-0 + 3-1 + 0-1 Summer ... .
+ 7-5 + 58-7 - 6-4 - 6-7 + 5-3 - 4-0 + 4-8 + 5-6 The differences in both oq and a2 in V also show a decided seasonal variation .
They are algebraically greater in summer than in winter ; but in the case of a4 the equinoctial values are the largest of all .
The variation of a4 in Y in the course of the year is only 42|- ' in ordinary as compared with 50|-0 in quiet days .
The reduction in the annual range of a2 in Y on ordinary as compared with quiet days is fully larger .
As regards a3 and a4 the differences between the ordinary and quiet day phase angles are small , and in the case of Y practically evanescent for the mean diurnal inequality of the year .
There is an irregularity in the incidence of the plus and minus signs in the case of the a4 differences which shows the advisability of reserve even as regards the seasonal figures .
In the case of a3 the plus and minus signs show a regular incidence , so that more reliance can be placed on the seasonal values .
It will be noticed that the seasonal variations in the two elements are exactly opposite .
YOL .
CCXVI.\#151 ; A. 2 M Before leaving the subject , attention may be called to the remarkable difference between the phenomena exhibited in the case of by the different elements .
In D the ordinary day phase angle substantially exceeded the quiet day angle throughout the whole year , the excess being greatest in the equinoctial and least in the summer months .
The excess in the mean diurnal inequality for the year represented about 29|- minutes* of time .
In H the ordinary day phase angle slightly exceeded the quiet day angle in summer , but fell notably short of it in winter , and in the diurnal inequality for the whole year was inferior to an extent representing about 6 minutes of time .
In V the ordinary day phase angle substantially exceeded the quiet day angle throughout the year , the excess being greatest in the equinoctial and least in the winter months .
In the mean diurnal inequality for the year the excess represented about 72 minutes of time .
Wolf 's Formula . .
S 23 .
If R denote the range of the mean diurnal inequality for the year , and S the sunspot frequency ( after Wolf and Wolfer ) , the formula R = a + b8 = a{l + ( fe/ a)S},.(l ) with a and b constants , was found by Prof. Wolf to represent closely the variation of the declination range with sunspot frequency .
It has been applied by myself to the ranges of the other magnetic elements , and to the individual months or seasons of the year as well as to the whole year .
In the case of the quiet day inequalities , the formula applied fairly to all the magnetic elements , being as closely true of H as of D. The value of b/ a was , however , not the same for the different elements .
Also , when the 12 months were treated separately , bja fluctuated from month to month , being greatest in winter and least in summer .
The formula has now been applied in the case of both H and V to the diurnal inequalities from ordinary days , for the year , the seasons , and the individual 12 months .
The results appear in Table XXXIII .
; lie is the unit in the case of a and b , and in the mean ( numerical ) difference between the values calculated by aid of the formula and those actually observed .
While a and b are given respectively to 2 and to 4 places of decimals , so as to show the exact values employed in the calculated ranges , the last figure possesses little if any physical significance .
Two sets of results are given for the year and the seasons .
In the first set the values assigned to a and b are arithmetic means from the individual months included .
The corresponding value given for b/ a is derived from these mean values of a and b , and is not the arithmetic mean of the values of b/ a for the individual months .
The second set of figures refers to the diurnal inequalities calculated for the year and seasons .
* There is a mistake in the description of Table XII .
of my previous paper dealing with the declination .
The data in it refer not to the seasonal inequalities , but to arithmetic means from the individual months comprised in the season .
FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
Taking , as an example , January in the case of H , the observed ranges in the Januarys of the 11 years were assumed to be given by a formula of type ( l ) , S representing the mean sunspot frequency for each January in succession .
The Table XXXIII.\#151 ; Constants in Wolf 's Formula II = Horizontal force .
Vertical force .
Mean difference Mean difference observed ^ observed ~ Month and calculated .
calculated .
Season .
a. b. 100 b/ a. Abso- As per- a. b. 100 b/ a. Abso- As perlute centage lute centage value .
OI value .
range .
range .
y y y y y y January 1112 0-1358 1-22 2-2 12-2 5-80 0-0666 1-15 0-8 13-6 February ... .
12-13 0-1495 1-23 1-5 11-4 9-43 0-1060 1*12 1-8 17-1 March 18-31 0-2674 1-46 2-4 12-9 19-26 0-1169 0-61 2-2 16-4 April 26-83 0-2946 1-10 3-0 12-6 25-45 0-0693 0-27 1-9 15-8 May 30-38 0-2232 0-74 3-8 11-7 27-74 0-1438 0-52 1-6 9-4 June 32-29 0:2561 0-79 3-0 12-1 25-12 0-1213 0-48 1-4 11-2 July 32-43 0-2450 0-75 4-3 14-6 24-56 0-1461 0-59 1-9 12-1 August 32-35 0-1825 0-56 4-2 16-7 23-28 0-0487 0-21 2-2 22-2 September . . . .
26-71 0-1831 0-69 3-1 16-4 16-73 0-1287 0-77 2-0 16*2 October 19-32 0-2635 1-36 2-2 11-2 12-16 0-1271 1-05 1-6 12-6 November . . . .
12-05 0-2195 1-82 2-3 12-3 8-63 0-0942 1-09 1-5 14-7 December ... .
9-67 0-1079 1-12 2-2 15-4 4-99 0-0712 1-43 1-2 12-6 From a.m.s. of - ' a 's and ft's\#151 ; Year 21-97 0*2107 0-96 2-8 13-3 16-93 0-1033 0-61 1-7 14*5 Winter 11-24 0-1532 1-36 2-0 12-8 7-21 0-0845 1-17 1-3 14-5 Equinox ... .
22-79 0-2522 1-11 2-7 13-3 18-40 0-1105 0-60 1-9 15-2 Summer ... .
31-86 0-2267 0-71 3-8 13-8 25-18 0-1150 0*46 1-8 13-7 From seasonal in- equalities\#151 ; Year 18-20 0-2081 1-14 0-5 3-3 16-17 0-1044 0-65 0-8 8-5 Winter 10-78 0-1544 1-43 0-7 6-1 6-79 0-0812 1-20 0-5 6-9 Equinox ... .
20-81 0-2588 1-24 1-2 6-6 17-87 0-1085 0-61 1-4 15-1 Summer ... .
30-95 0-2417 0-78 1-5 6-1 24-58 0-1198 0-49 1-3 9-9 appropriate values of a and b were then determined by least squares , and proved to be a = 11'12-y , b = 0'1358y .
Inserting these values of a and b in the formula , and ascribing to S in succession the frequencies of the 11 Januarys , values were deduced for II .
The observed values of H in January fluctuated from 8'ly in 1899 to 25'8y in 1892 , i.e. , they had a range of 17'7y .
The sum of the eleven differences between observed and calculated values taken irrespective of sign was 23'7y , the mean difference being thus 237/ 11 or 2'2-y .
This expressed as a percentage of the range is 237 x 100/ 11 x 177 , or 12'2 .
There is obviously a good deal that is " accidental " in the results obtained for individual months , especially in Y. The values of it will be seen , are decidedly larger for H than for V , implying that relatively considered the sunspot influence has more effect on the range of the former element than on that of the latter .
The differences between observed and calculated values naturally tend to increase in size with the extent of the variation of R during the 11 years , and so are largest in summer .
The fairest way , however , to compare the accuracy of the formula at different seasons , or for different elements , is to take as criterion the percentage which the mean difference between observed and calculated values is of the difference between the largest and least observed values of R. In H these percentages show no marked variation with the season .
There are considerably greater variations in Y , but they probably represent in the main the smaller accuracy of observed values in that element , as the largest and smallest values for individual months both fall in summer .
If we exclude the equinoctial value in Y , the agreement between observed and calculated values is decidedly closer for the seasons than for individual months , and in the case of H it is still closer , to a marked degree , for the year .
One 's estimate of the suitability of a formula , while mainly determined by the size of the mean difference between observed and calculated values , is partly determined by the mode of grouping of the plus and minus signs .
For instance , if the plus signs all occurred together , there would be reason to suspect a sensible secular change in the amplitude of the diurnal inequality , quite apart from sunspot variation .
Table XXXIY .
accordingly records for each year the actual differences between the observed Table XXXIY.\#151 ; Observed less Calculated Ranges in mean Diurnal Inequality for the Year .
Element .
1890 .
1891 .
1892 .
1893 .
1894 .
1895 .
r 1896 .
1897 .
1898 .
1899 .
1900 .
H + 0* lie + 0 .
6y + 0 ' 5y \#151 ; 0-3y + 0 * 4y \#151 ; 0.5y \#151 ; 0 ' lie 0'0y ' -l-4y + 1 * 3y - 0 * 6y y -l-7y + 1 * 2y 0-0y \#151 ; 1 .
6y 4- 0 ' 4y 0 * Oy + l-0y + 0 * 8y +1 .
lie 0*0y - 1 * lie i + 0 ' .
05 + 0 ' .
05 + 0'-08 + 0'-01 O'-00 - O'*06 - O'-07 - O'*04 - O ' -13 + 0'*13 !
- O'*03 and calculated values .
The calculated values for II and V were derived from the values of a and b assigned in Table XXXIII .
to the mean diurnal inequality for the year .
The results for I were calculated from a = Cf'8648 , = O'-Ollll , these being the values obtained by the aid of least squares from the observed I range , in the mean diurnal inequalities for the eleven years .
The corresponding value of FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
255 100\amp ; /a , viz. , 1*28 , it may be pointed out , is considerably in excess even of that found for H , and double that found for V. An analogous result was observed in the case of the quiet day ranges .
The plus and minus signs occur pretty promiscuously in Table XXXIY .
Also there seems no parallelism between the differences observed in the case of the different elements .
In short the differences do not suggest anything but accident .
Considering that the observed H ranges varied from 19'6y to 35'67 , it seems not a little remarkable that in seven years out of the eleven , the difference between the observed and calculated value did not exceed 0'5y .
The mean difference between observed and calculated values of I was 0'*059 , or 67 per cent , of the range in I during the eleven years .
The agreement is thus closer than in the case of Y , though decidedly less good than in the case of H. S 24 .
Wolf 's formula was also applied to the observed amplitudes of the 24- , 12- , 8-and 6-hour terms in the mean diurnal inequality for the whole year , in the case of H and V. The results appear in Table XXXY .
The agreement between the observed and calculated values is in general good , especially in the case of the 24- and 12-hour Table XXXY.\#151 ; Constants in Wolf 's Formula K = a + Z\gt ; S. Horizontal force .
V ertical force .
\#151 ; a. 1 6 .
1006/ a. Mean difference ol J*er vod ~ calc ulated .
a. 6 .
100 b/ a. Mean difference observed-^ calculated .
Absolute value .
As percentage of range .
Absolute value .
As percentage of range .
Cl 7 6-32 7 0-0868 1-37 7 0-27 4-1 7 5-62 7 w 0-0317 0-56 7 0-40 12-1 C2 3-49 0-0440 1-26 0-21 6-0 3-42 0-0260 0-76 0-15 7-0 c3 2-04 0-0103 0-51 .
0-14 10-5 1-32 0-0090 0-68 o-io 10-0 Cl 1-05 0-0028 0-27 0-04 12-1 0-49 0-0023 0-47 0-05 22-7 terms in H and the 12-hour term in V. It is least good in the case of the 6-hour terms , whose amplitudes even at their largest are very small .
The general tendency apparently is for b/ a to become smaller , i.e. , for the sunspot influence to be less marked , as the order of the harmonic , increases , but the 24-hour term in V seems exceptional .
The values obtained for a , b , and b/ a for ordinary days in the case of H are very similar to those obtained from the quiet day ranges , but are on the whole slightly larger .
In the case of V the differences between ordinary and quiet day results are more marked , and the excess in the b/ a from ordinary days over that for quiet days is larger .
Daily Range .
S 25 .
The term daily or diurnal range is used in several senses .
It may mean the difference It between the algebraically largest and least hourly values in the diurnal inequality , that inequality being derived from selected days or from all days .
It may mean , however , the difference It ' between the highest and lowest daily values , irrespective of the time at which they occur , whether an exact hour or not .
The mean It ' for a month is simply the arithmetic mean of the values for individual days of the month .
Tables XXXVI .
and XXXVII .
give It for H and V from ordinary days , for each month of the eleven years .
Tables XXXVIII .
and XXXIX .
give It ' from all days for the same two elements .
The It derived from any combination of days must be less than the corresponding II ' , unless the maximum and minimum each occur at a fixed time , which is an exact hour .
In practice , the times of the maximum and minimum vary from day to day .
Speaking generally , the difference between It and R ' for a particular month is larger the more disturbed the month .
If instead of the month one takes the year , the mean R ' is the arithmetic mean of the values for the 12 months , but R is less than the arithmetic mean of the monthly values unless the hours of maximum and minimum in the inequality are the same throughout the year , which is never the case at Kew .
Even when one considers the months of the same name from a number of years , there is usually some variation in the hour of maximum or minimum .
To bring out these points , Tables XXXVI .
and XXXVII .
contain the values of R from the inequalities Tor the whole year and for all the months of the same name combined , as well as the arithmetic means of the values for the individual months .
In Table XXXVI .
, November is the only month in which the R from the months combined equals the arithmetic mean of the values for the separate months .
In the other 11 months the former quantity is the smaller , though the difference is never large .
There is a much larger difference between the inequality range for the whole year and the arithmetic mean of the values of R for the twelve months .
The former quantity on the average stands to the latter approximately in the ratio 7 .
8 .
In Table XXXVII .
, April is the only month in which R from all the months combined is the same as the arithmetic mean of the values for the individual months , but the differences in the other months are small , just as in the case of H. The differences in Table XXXVII .
between the inequality range for the year and the arithmetic mean of the ranges for the twelve months are a good deal smaller than in the case of H. The average excess of the arithmetic means is only about 3^ per cent. This implies , of course , less variability in the hours of maximum and minimum .
Comparing Tables XXXVIII .
and XXXIX .
with Tables XXXVI .
and XXXVII .
we see that the excess of R ' over R is usually large .
Allowance must , of course , be made for the fact that the more highly disturbed days\#151 ; 242 for H and 348 for V\#151 ; Table XXXVI.\#151 ; Horizontal Force .
All Ordinary Days .
Inequality Ranges .
( Unit Lie .
) FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
\TB9if JOJ ifyjl'enbouj OOiMCiOOOOOOO^O^ OiCDCOiO^H ?
DCON(M05 h(NCOCOWW^(N(N(Nh 26-9 .ireaui opeimpuy CiOOiO5OOO0OO5(N75 WONOJOJiOONiOO M ( MCOCOCOCOCOCO(M(M(M(M 30-7 *JOqiU0O9Q Oi(MCOOl\gt ; L\gt ; OOTH01NlO COO^WOOCOO^tMOiOOO i-H i-H i-H ( M r-H i-H i-H i-H 14-0 13-6 ' M9qiH9AO^J CD(MCO(M^(MO^OONCO I-H ( M CO CO CO CO r\#151 ; 1 i-H i-H I\#151 ; H I-H 19-7 19-7 .agqo^OQ COOCONOOHO(NJ\gt ; ( M^ COOOHONCOlO^CO(MCO coco^^cocococ^cococo 30-3 29-9 .J9qui9^d9g CO(MCOOOC^C5tOOOCOON OOiOCOHCOOCOHTt^ COCOCO^^CO^COCOCOCO 35-0 34-5 .^snSny CJO^CiWIOCOHNOOCO ihCOiOhCIN(MCOtH^O co^ioioioco^cocococo 40-6 40-2 .Anr OOlOONOOON W'tOONOClOOQCDOO CO^iOiOiOiO^COCOCO(M 43-4 430 *9unf COOO^CDCiOCJNCJHH cocrsoot^t^\lt ; x\gt ; GOt^r^ooco COCO^iOiOROCOCOhHCOCO 43-9 43-8 O lO H o O \lt ; N D CO O N H IC^IOOOOLO OO^ONO COCO^^HlO^^^COCOCO 40-0 39-9 .Iiady NiOiOOONlONClrHOO OOOiO^OJO^CICDOOCD 39 0 38-3 .qoj'Bpj CO N CDO 0 O 00 Cl ( N 00 lO OCUOtOHQOOilOC'lCiCC CO CO CO CO CO CO GO CO CO r-H CO 27-8 26-7 .^iRiuqgj CJOHOO^OlO^fNCOCO Tt \#187 ; c Cl tH CO 1\#151 ; it\#151 ; CO^CO i-H r-H CO CO CO CO CO r\#151 ; 1 r-H i-H i\#151 ; 1 18-5 180 Q H GO QOl\gt ; H 1\#151 ; i 05 Cl 1\#151 ; ILO HCOiQl\gt ; HOOCll\gt ; ^QOCO i\#151 ; H i\#151 ; H CO i\#151 ; H CO i-H i\#151 ; H i\#151 ; H i\#151 ; H l-H 16-5 16-0 Year .
OH(NC0^\#187 ; 0Ol\gt ; 00C5O 05050)010^0505050500 OOOOOOCOOOOOOOQOOOOOC5 I-H I-H I-H I-H I-H I-H I-H r-H t-H i\#151 ; H r-H f Arithmetic 1 mean . .
j 11-year in-1 equality .
j \lt ; D bJD CC \lt ; D I\#151 ; I CD O hi O EH .S '-S hi \lt ; D hH hH t\gt ; X H x H iJ fa \lt ; 1 Eh IOJ ^ijHnbauj ( MHQOhhOOlOhHhH lQHOOW4(NHQOhO H(M(M(M(M(N(NH^HH 20-5 *ni30ni 1 : ^COHOO(M(M1\gt ; ( MOO \lt ; M oporaq^ijy \#187 ; OHHHiOCO(MOOOO0 r\#151 ; i \lt ; M CM .cm NN(MN(NHH r-H I \lt ; M h O h Q O 4 h ( M O C5 lO CCi i\#151 ; H .J0qiU000Q COOO(MCiOOhiO01\gt ; iO r\#151 ; H i-H .
i-H I\#151 ; CDNHCDOO(M^OOOOhO a* io *j0qui0AO^[ \gt ; QCO(MiOiOiOOOh005 r-H r-H r-H r-H r\#151 ; 1 r-H rH r-H i\#151 ; H r\#151 ; H r-H r-H i\#151 ; H ( M(NOOJ !
NO(MOhiOGO io o \10qocpo OiH(MOHOl^OO(MiO r-H rH ( M !
M(M(M(Mhhhhh O co 00*00 IO N h CO CO CD H IO t-H .J0qraoad0g NC0C0C000G04(M4OC0 \lt ; M \lt ; M \lt ; M \lt ; M h CM \lt ; M \lt ; M \lt ; M H CM \lt ; M \lt ; M \lt ; M H OOOtMOOOOOCDOhCOCO lO co .^snSny OOOOOOb\#187 ; OOiOhCOCOO(M lO i\lt ; 0 \lt ; M \lt ; M i\#151 ; I\lt ; M\lt ; M\lt ; M\lt ; M\lt ; M\lt ; M\lt ; M\lt ; M\lt ; M\lt ; M 'Anr COOW(NOOCOCDOOhOCO r-H OO iO H CD Oi CD CD O lO GO CD CO i-H O co co ( M CO CO CO CO CO CO ( M ( M Cl ( M ( MCMi-HCOr-HfMOOOOi-HOi OO -4 .0unf 4OC0CDCDCDH00OiO\#187 ; 0 o o CO co ( MCOCOCOCOCOCO(MCO(M(M iOOOCOH(MOOiiOCOiOOi CO OO OdOhtMOHCOCOOJh co co co co OICOHCOHCOCOCOCOOKM .[udy ( MCDOOCDCMOCOCiOOlOOl co co HOOhOCOOOJOihOOH GO 00 CM\lt ; M\lt ; MCOCOCO\lt ; M\lt ; M\lt ; M\lt ; M\lt ; M \lt ; M \lt ; M OhCOOcDCDOOHOiCDn -4 O OiCOO\lt ; M4cOhCOHOh CO co H(MCO(M(M\lt ; M(M\lt ; M\lt ; M\lt ; Mh \lt ; M \lt ; M COCOcDCOOlOiOhCOCDH O 00 *^jmaq0lq HOOCDHOOOOhhiOQQ -h CO r-H rH i-H r-H t-H i-H i-H rH i-H i-H l\gt ; \lt ; MJ\gt ; CiCO\lt ; M\lt ; MlO\lt ; MOOCl 'h cn .XmuiT ?
jp UOOrHOOOOt- \lt ; \#163 ; \gt ; \lt ; \#163 ; \gt ; *- r-H r-H r-H i-h r-H 00 iH o3 \lt ; D OHtMCOhiOCDhOOOiO 0(j505C5QOC5(J5C5CCiO GO OO GO OO GO OO 00 CO CO GO C5 Arithmetic 1 mean .
J 11-year in-1 equality .
j \gt ; * i\#151 ; H i\#151 ; H i\#151 ; H i\#151 ; H i\#151 ; H i-H i\#151 ; H r-H r\#151 ; H j\#151 ; H r\#151 ; H FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
contributed to R ' but not to R. If they had been excluded in both cases , R ' would have been appreciably reduced , more especially in the equinoctial months and in the Table XXXYIII.\#151 ; Horizontal Force .
All Days .
Absolute Ranges .
( Unit lie .
) Year .
# J anuary .
February .
March .
April .
May .
June .
July .
August .
September .
October .
November .
December .
Mean .
1890 35-1 43-1 39-6 46-3 40-3 47-4 50-5 52-6 55-8 54-1 41-8 31-4 44-8 1891 34-9 47-4 66-8 74-6 75-9 60-0 61-4 69-3 74-3 71-4 50-7 43-0 60-8 1892 56-2 113-0 114-1 83-9 91-2 83-1 117-3 96-1 70-6 74-6 54-9 53-4 84-0 1893 55-4 59-1 67-6 77-5 65-3 85-7 83-1 88-5 79-8 71-5 60-4 43-1 69-7 1894 61-6 98-5 78-3 78-8 82-8 95-0 106-4 95-1 88-6 71-9 74-2 46-2 81-4 1895 47-4 66-0 75-6 81-3 76-2 82-3 80-5 58-0 62-7 73-3 64-4 46-5 67-9 1896 63-5 68-2 68-5 74-4 78-7 60-5 68-0 75-7 76-9 58-1 43-1 38-9 64-5 1897 40-0 41-8 51-8 77-6 68-9 55-6 53-9 51-2 48-3 50-7 37-8 44-1 51-8 1898 39-4 49-0 73-1 51-7 64-6 62-3 60-3 61-7 77-0 54-8 40-0 37-8 56-0 1899 39-3 42-1 50-4 55-9 64-7 60-5 55-0 57-1 59-9 41-7 33-6 35-0 49-6 1900 39-1 33-4 47-8 41-4 48-6 43-5 40-2 42-6 36-1 34-7 22-7 18-1 37-3 Mean .
46-5 60 -T 66-7 67-6 68-8 66-9 70-6 68-0 66-4 59-7 47-6 39-8 60-7 years 1892 , 1894 , and 1896 ; while , if they had been included in both cases , R would probably have been slightly increased .
Table XXXIX.\#151 ; Vertical Force .
All Days .
Absolute Ranges .
( Unit lie .
) Year .
January .
February .
March .
April .
tA cS S June .
July .
August .
September .
October .
November .
December .
Mean .
1890 16-6 20-6 26-5 27-2 28-4 26-8 25-8 23-0 21-9 22-0 15*5 3-6 21-5 1891 15-5 23-9 35-9 42-8 53-7 31-6 35-8 33-5 47-3 32-2 25*5 24-8 33-5 1892 25-0 65-8 74-9 49-0 68-4 52-2 68-0 48-3 33-0 39-9 28*3 32-8 48-8 1893 27-0 33-1 34-1 35-3 38-6 44-0 44-7 46-9 39-9 34-6 37*4 24-4 36-7 1894 28-8 68-6 52-8 45-7 44-9 45-9 65-8 48-7 58-0 29-1 45*8 21-8 46-3 1895 23-8 38-8 38-9 40-8 44-0 43-3 38-6 30-8 30-7 40-6 32*6 24-0 35-6 1896 28-8 34-7 44-8 37-8 51-7 32-4 36-2 37-2 32-7 30-0 22*6 22-3 34-3 1897 23-5 21-2 28-9 44-1 37-9 29-4 28-0 27-8 23-8 20-7 16-6 19-3 26-8 1898 17-2 26-1 53-3 30-8 36-4 30-8 29-3 33-0 44-1 27-6 22*9 18-0 30-8 1899 21-1 25-5 29-2 31-7 33-0 33-2 29-2 24-6 24-3 17-7 14*9 17-4 25-1 1900 19-0 16-8 27-0 23-4 34-8 26-8 25-8 26-0 19-3 19-7 13*9 12-8 22-1 Mean 22*4 34-1 40-6 37 1 42*9 36-0 38*8 34*5 34*1 28*6 25-1 20*1 32-9 VOL. ccxvi.\#151 ; A. S 26 .
Table XL .
gives the ratio borne by the mean B/ to the corresponding arithmetic mean of the values of It for the 12 months and for the 11 years separately .
The disturbed days omitted from the V inequalities were the more numerous , so that any reduction in R consequent on this exclusion would naturally be greater in Y than in H. The excess of R ' is , however , invariably much larger for H than for V , confirmatory of what has been already said as to the former being in general much the more disturbed element .
Relatively considered , the excess of R ' over R has a conspicuous maximum near mid-winter , and a minimum near mid-summer .
The pre-eminence of the values of R'/ R in winter over those in equinoctial months is hardly what one would have expected from consideration of disturbances alone .
We see that the size of the regular diurnal inequality is apt to give in winter a very inadequate idea of the average diurnal variation in the field .
Table XL.\#151 ; Values of ( Absolute Range from All Days)/ ( Inequality Range from Ordinary Days ) .
Element .
January .
February .
March .
April .
May .
June .
July .
August .
September .
October .
November .
December .
Mean .
h 2-81 3-25 2-40 1-73 1-72 1-52 1-63 1-67 1-90 1-97 2-41 2-84 2-13 y 2-65 2-44 1-73 1-31 1-26 1-18 1-25 1-35 1-51 1-64 2-10 2-56 1-75 1890 .
1891 .
1892 .
' 1893 .
1894 .
1895 .
1896 .
1897 .
1898 .
1899 .
1900 .
Mean .
H 1-90 1*99 2-22 1*74 2-09 1-90 2-11 1*88 2*16 1-97 1-67 1*97 V 1-37 1*54 2-00 1-49 1-82 1-49 1-55 1*32 1-49 1-38 1-32 1-52 If we compare the values of R'/ R in different years , we see that while on the whole it was least in the years of fewest sunspots , it was considerably below average in 1893 , the year of sunspot maximum .
In the case of H , in fact , it was lower for 1893 than for any other year except 1900 .
This means that in 1893 R was specially large , and not that R ' was small .
As Table XXXVIII .
shows , the mean value of R ' for 1893 was 15 per cent , above the average of the 11 years , and was exceeded only in 1892 and 1894 .
In the case of V the value of R/ /R for 1893 , though less remarkable , was below the mean .
In both H and V the values of R ' for 1892 and 1894 were much in excess of that for 1893 .
For 1892 this excess was 21 per cent , for H , and 33 per cent , for V. Table XLI .
shows the order in which the years stand when placed in descending order of range .
Data for D are included to make the survey more complete .
The FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
resulting order is also shown when the figures from D , H and V are added , , and finally the order when the sunspot frequency is the criterion .
R and R ' are treated separately .
When the size of R ' is the criterion , the order is the same for all the magnetic elements , except that 1890 comes last in V , while 1900 comes last in D and H. There is less agreement between the results from the three elements in the case of R , but the differences between the pairs of years 1890 and 1900 , 1892 and 1894 , and last 1897 and 1898 , are so small that little significance attaches to the precise order in which the members of the pair present themselves .
In the final returns from the Table XLI.\#151 ; Position of Years when Arranged in Descending Order of R , R ' , and S. 1890 .
1891 .
1892 .
1893 .
1894 .
1895 .
1896 .
1897 .
1898 .
1899 .
1900 .
D 9 6 2 1 3 4 5 7 8 10 11 R II 10 6 3 1 2 4 5 7 8 9 11 V 11 6 3 2 1 4 5 8 7 9 10 Mean . .
10 6 3 1 2 4 5 7 8 9 11 D 10 6 1 3 2 4 5 8 7 9 11 R ' H 10 6 1 3 2 4 5 8 7 9 11 V 11 6 1 3 2 4 5 8 7 9 10 Mean . .
10 6 1 3 2 4 5 8 7 9 11 Sunspots ... .
11 6 3 1 2 4 5 8 .
7 9 10 three elements combined , the only difference between the R and R ' lists is that there is an interchange of place between 1892 and 1893 , and between 1897 and 1898 .
Table XLII .
deals with the twelve months in the same way that Table XLI .
dealt with the eleven years .
In Table XLII .
the points of agreement between the different elements in the same list , or between the same element in the two lists , are somewhat inconspicuous outside the mid-winter months .
Successive months sometimes occupy very different positions , without there being a very large difference between their values of the ranges .
A longer period of years would be required to eliminate satisfactorily what is accidental .
Whether one takes R or R ' , it is clear that December , January , and November fill the last three places .
When the three elements are combined , May , July , and August are bracketed second in the case of R. In the case of R/ there is a bracket between April and July , and again between June and September .
2G2 DR. C. CHREE : DISCUSSION OF KEW MAGNETIC DATA The difference between the positions occupied by June in the It and R ' lists is remarkable .
In the It list it comes first , slightly in advance of the other three midsummer months , whereas in the It ' list no element places it higher than fifth .
Another striking phenomenon is the difference between the positions assigned to March and July in the It ' list according as the element considered is D or H. In the case of H , it is true , the mean values of It ' for the six months March to September differ comparatively little .
Still the fact that ( July Range/ March Range ) = 0'88 for D , but = 1'06 for H , appears a little remarkable .
Table XLII.\#151 ; -Position of Months when Arranged in Descending Order of R and R ' .
January .
February .
March .
April .
1 May .
June .
July .
August .
September .
October .
November .
December .
D 11 9 7 2 4 3 5 1 6 8 10 12 R H 11 10 8 5 4 1 2 3 6 7 9 12 Y ii 9 6 4 1 3 2 5 7 8 10 12 Mean .
11 9 7 5 ( 2 ) 1 ( 2 ) ( 2 ) 6 8 10 12 D 11 8 1 2 3 9 6 5 4 7 10 12 R ' H 11 8 6 4 2 5 1 3 7 9 10 12 Y 11 7 2 4 1 5 3 6 8 9 10 12 Mean .
11 8 2 ( 3 ) i ( 6 ) ( 3 ) 5 ( 6 ) 9 10 12 S 27 .
In the case of the mean diurnal inequality for the year there is , as we have seen , a close connection between the range and the corresponding sunspot frequency S. This does not necessarily imply any close connection between sunspot frequency on any given day and the amplitude of the regular or irregular magnetic changes on the same day .
It seems increasingly difficult to think of any cause for the magnetic diurnal inequality other than electrical currents in the upper atmosphere .
In temperate latitudes , whether at sunspot maximum or minimum , regular magnetic changes are most rapid during the day , but the difference between day and night seems reduced at sunspot maximum .
Whether by day or by night , the regular changes are larger in sunspot maximum than in sunspot minimum .
The most natural inference is that direct sunlight , whether there are or are not sunspots , increases the conductivity of the upper atmosphere , and that the effect persists to some extent during the night .
At sunspot maximum the upper atmosphere is more conducting than at sunspot minimum at the same hour .
The state of the upper atmosphere as regards FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
conductivity may be due to contributions from some , perhaps many , previous days .
The solar influences may be of different kinds , taking different times to travel from the sun and decaying at different rates .
Sunspots again maybe evidence of some effect on the sun which is shared by the solar system , but which takes some time to travel the distance separating the earth from the sun .
The diurnal inequality can be derived only from a combination of days .
Thus it does not enable us to compare magnetic conditions and sunspot frequency on individual days .
There is even a difficulty in comparing the run of magnetic conditions and sunspots during successive months , owing to the annual variation in the daily range .
This difficulty can , however , be fairly surmounted if we express the range for each month as a percentage of the mean range from all months of the same name in the 11-year period .
The percentage values thus obtained for R and R ' in H appear in Table XLIII .
along with the corresponding Wolfer 's sunspot frequency , the latter in heavy type .
The general tendency for the percentage figures in Table XLIII .
to be large in years of many , and small in years of few sunspots , is of course obvious .
But when we compare successive months , we see that with rise of S we may have rise or fall of R , and that R and R/ not infrequently change in opposite directions .
There being 132 months , there are 131 passages from one month to the next .
If we allow 1/ 2 in cases where the value is the same in two consecutive months , we find that R and R ' changed in the same direction in 84 cases , S and R changed in the same direction in 65^- cases , while S and R ' changed in the same direction in 72|- cases .
A large monthly value for R ' may be due to only two or three highly disturbed days , excluded from the ordinary days , still the number of cases in which R and R/ changed in opposite direction is larger than would have been expected .
The figures quoted above , by themselves , afford no evidence of a connection between S and R in individual months , and only slight evidence of a connection between S and R ' .
This differs from what was observed in the case of D. There R and S increased or diminished together in 75 cases out of 131 , while R'and S increased or diminished together only in 68 cases .
In many instances the values of S for consecutive months differ so little that accident might play a considerable part .
If we confine ourselves to the 52 cases in which S changed by at least 10 units , there was agreement in the direction of change in 30 cases as between R and S , This is decidedly more favourable to a connection between R and S in individual months .
In a second investigation the months of each year were arranged in two groups of 6 , consisting respectively of the months of largest and least sunspot frequency .
The Table XLIII.\#151 ; Comparison of Sunspot Frequencies and H Range Percentages .
Year .
b c3 P C \gt ; \gt ; \lt ; 8 B \#166 ; g A g j-H Ph J3 \#169 ; P P r\#151 ; H 4-j GO CJ \lt ; D rP a \#169 ; Ph \#169 ; 1 S3 \#166 ; S U \#169 ; r-Q a \#169 ; \gt ; 0 Fh \#169 ; rP a \#169 ; \#169 ; \#169 ; \#166 ; \lt ; s m O \#163 ; s 5- 3 0-6 5 i 1 6 4* 8 1 .3 ii- 6 8 .5 17 ' 2 11-2 9 .6 7 .8 1890 R 72 80 74 73 64 76 77 79 87 77 79 99 R ' 75 72 59 68 59 71 * 72 77 84 91 88 79 S 13- 5 222 10- 4 20- 5 41- 1 48* 3 58- 8 33 2 53 8 51-5 41 9 32 \#166 ; 2 1891 R 79 76 82 91 86 91 102 106 112 125 133 116 R ' 75 79 100 110 110 90 '87 102 112 120 107 108 S 69- 1 75-6 49- 9 69- 6 79- 6 76 .
3 76- 8 101- 4 62 ' 8 705 65 \#166 ; 4 78 .6 1892 R 157 136 127 116 113 110 134 136 104 137 123 102 R ' 121 188 171 124 133 124 166 141 106 125 115 134 S 75- 0 730 65- 7 88- 1 84- 7 88- 2 88- 8 129 ' 2 77 .
9 797 75 1 93 \#166 ; 8 1893 R 108 123 131 130 116 131 132 128 125 134 153 164 R ' 119 98 101 115 95 128 118 130 120 120 127 108 S 83- 2 84-6 52- 3 81- 6 101 .
2 98- 9 106- 0 70- 3 65 ' 9 75-5 56 6 60 0 1894 R 131 134 114 127 145 132 117 129 118 125 119 134 R ' 132 164 117 117 120 142 151 140 133 120 156 116 S 63- 3 672 61- 0 76- 9 67- 5 71- 5 47- 8 68- 9 57- 7 67-9 47- 2 70 ' 7 1895 R 110 124 137 130 113 129 121 92 97 109 123 119 R ' 102 110 113 120 111 123 114 85 94 123 135 117 S 29- 0 57-4 52- 0 43- 8 27- 7 49- 0 45- 0 27- 2 61- 3 284 38- 0 42- 6 1896 R 116 116 107 114 102 89 92 104 115 84 76 70 R ' 137 114 103 110 114 90 96 111 116 97 90 98 s 40- 6 29*4 29- 1 31* 0 20* 0 I4- 3 27- 6 21- 8 48- 1 14-3 8 ' 4 33- 3 1897 R 108 95 93 109 102 86 85 82 82 80 63 89 R ' 86 70 78 115 100 83 76 75 73 85 79 111 S 30 ' 2 364 38- 3 14- 5 25- 8 22- 3 9- 0 31- 4 34- 8 344 30- 9 12- 6 1898 R 90 66 80 69 90 95 90 85 89 78 90 71 3 ' 85 81 110 77 94 93 85 91 116 92 84 95 S 19 5 9*2 18 1 14 2 7- 7 20- 5 13- 5 2- 9 8- 4 130 7- 8 io- 5 1899 R 49 79 71 72 94 87 85 86 100 73 85 76 R ' 85 70 76 83 94 90 78 84 90 70 71 88 S 9 .4 136 8 .6 16* 0 15* 2 12- 1 8- 3 4- 3 8- 3 12-9 4* 5 o- 3 1900 R 82 72 85 69 75 75 66 75 71 77 57 61 R ' 84 56 72 61 71 65 57 63 54 58 48 46 FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
mean values obtained for S and for the percentage values of R and R ' in H for the two groups appear in Table XLIY .
Table XLIY.\#151 ; Percentage H Ranges from Groups of Months of Larger and Smaller Sunspot Frequency .
Year .
Group of m onths of largest sunspot frequency .
Group of months of least sunspot frequency .
s.- R per cent. R ' per cent. s. R per cent. R ' per cent. 1890 11-0 83 82 3-1 73 67 1891 49-2* 108 104 22-0 92 96 1892 81-4 122 148 64-6 127 127 1893 95-5 133 116 74-4 129 114 1894 92-6 131 138 63-4 123 130 1895 70-6 115 113 57-4 119 111 1896 51-4 106 105 32-2 92 108 1897 35-3 96 89 17-2 83 83 1898 34-4 81 96 19-1 84 88 1899 16-5 73 80 7-7 87 83 1900 13-2 75 66 5-7 69 57 Mean . . .
50-1 102-1 103-4 33-3 98-0 96-7 The higher R ' percentage appears in the group of higher values of S in 9 years out of the 11 .
In the two exceptional years , 1896 and 1899 , the deficiency in the value of R ' in the first group is only 3 , and so possesses little significance .
On the average of the eleven years , the excess of the percentage values of IF in the first group amounts to 67 .
If we regard this as a percentage on 607y , the mean value of R in H for the eleven years , we get 4'07y as corresponding to a'difference of 16'8 in S. On a formula of the type IF = a + b S , this would give \#166 ; 100 b/ a 0-47 .
In the case of R the higher percentage , value is associated with the higher vahie of S in 7 years .
In one of the 4 exceptional years , 1899 , the deficiency in R is substantial , but this possesses less significance than would otherwise be the case from the fact that the difference between the values of S for the two groups in that year is little over half the average .
The excess in the percentage value of R in the group of higher values of S , on the average of the 11 years , is 4 1 .
If we regard this as a percentage on 307y , the mean value of R in H from the 132 months , we get l'26y as corresponding to a difference of 16'8 in S. On a formula of the type R = S , this would give 100 bja = 0'27 .
Taking the arithmetic mean of the monthly values of It in H for each year of the eleven , and applying the It = a + bS formula , I found by least squares a = 21'47y , b = 0'2224y , 100 6/ a = 1'036 .
The division of the year into equal groups of 6 months only partly attains the object in view , viz. , the elimination of any cause whose effect lasts for several months , because there is some tendency for the months of higher S in a year to occur together .
In 1890 , for instance , the 6 months of largest S came from the second half of the year , while in 1894 and 1900 the first half of the year contributed 5 out of the 6 months in the first group .
Taking this into account , and the smallness of the value of 6/ a deduced in the case of It from Table XLIV .
, the natural inference would seem to be that the value of It from ordinary days in a particular month depends less on the value of S in that month than on the values of S in the other months of the year .
This may mean nothing more than that the value of S on a particular day is a very rough measure of the solar activity to which enhanced diurnal inequalities are due , while the mean value of S from all days of the year affords a very exact measure .
The fact that the value of b/ a derived from Table XLIY .
in the case of IF is so much larger than that derived in the case of It , suggests that the solar influence is more direct or immediate in the case of irregular than in the case of regular magnetic changes .
Table XLY .
shows the result of grouping the 132 months according to the monthly mean value of S. Notwithstanding the considerable number of months included in the groups , the results of the first grouping in which the step in S is only 10 are somewhat irregular .
It and IF show a decided tendency throughout to increase with Table XLY.\#151 ; Percentage Ranges from Months Grouped according to Sunspot Frequency .
Range of values of S. Months in group .
Mean values .
Range of values of S. Months in group .
Mean values .
s. R per cent. R ' per cent. s. R per cent. R ' per cent. 0 to 10 23 6-2 77-8 71-5 0 to 20 43 9*6 76-6 74-4 10 " 20 20 13-8 75-1 77-7 20 " 40 26 29*1 91-4 96-5 20 " 30 14 25-1 93-0 94-6 40 " 60 20 49*3 106-4 109-2 30 " 40 12 33-7 89-5 98-7 60 " 80 30 70*6 123-0 121-7 40 " 50 12 45-4 103-0 106-7 \gt ; 80 13 94*6 132-5 127-8 50 " 60 8 55-0 '111-5 112-9 60 " 70 14 65-4 120-2 112-6 70 " 80 16 75-2 125-4 129-6 80 " 90 7 85-6 128-7 124-1 \gt ; 90 6 105-1 137-0 132-0 FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
S , but in the group or groups with S above 80 the rate of increase seems reduced .
The exceptionally high monthly values of S occurred sporadically .
For instance , the extreme value , 129'2 in August 1893 , was preceded by 88'8 in July and followed by 77'9 in September .
Thus we should not expect a correspondingly high value of It unless the influence of previous months is negligible .
In considering the significance of Table XLV .
allowance should be made for the fact that some of the groups come entirely from years of sunspot maximum .
No value of S as high as 60 occurred in the sunspot minimum or intermediate years .
Again , sunspot minimum years gave no value of S larger than 20'5 , and so their contributions were practically confined to the groups 0 to 10 and 10 to 20 .
Intermediate years contributed only two months to the group 0 to 10 , and six to the group 10 to 20 .
S 28 .
We have hitherto considered only mean monthly values of B/ , but the way in which these values are made up seems also of interest .
To make the results more complete , D has been included in the investigation .
Some data for D were got out in a previous paper , * but they are considerably extended here .
As a preliminaiy it is well to know something about the extreme values of I !
/ .
To this end Tables XL VI .
and XLVII .
give the smallest recorded values of B/ in D and H for each month .
Table XLVI.\#151 ; Declination .
Absolute Daily Banges .
Smallest Values .
Year .
January .
February .
March .
April .
May .
June .
July .
August .
September .
October .
November .
December .
Year .
1890 4-2 / 5-6 / 5-1 / 8-2 6-2 6-9 / 8-0 / 8-2 5-0 5-7 r 5-0 / 3-8 / 3-8 1891 31 4-1 4-3 8-1 8-5 7-9 8-6 9-2 9-8 7-9 7-1 4-7 3-1 1892 5-7 8-3 11-4 10-9 10-3 9-2 11-4 10-1 9-2 8-5 3-7 5-7 3-7 1893 5-0 5-0 8-0 12-5 11-9 12-3 11-6 11 -5 10-1 9-3 5-9 4-1 41 1894 6-8 7-9 8-3 9-6 10-6 10-5 6-8 9-7 10-3 9-0 4-7 4*6 4-6 1895 3-9 6-8 8-3 11-9 10-0 13-0 10-8 8-0 8-5 8-6 5-2 3-8 3-8 1896 6-0 7-3 8-0 10-9 9-0 6-7 7-9 9-8 8-2 6-6 34 2-7 2-7 1897 3-2 4-3 8-9 8-3 7-4 7*8 8-0 9-1 6-1 5-8 4-3 2-2 2-2 1898 3-1 3-5 6-0 8-0 7-1 7-7 7-2 8-0 7-5 6-5 4-3 33 3-1 1899 3-0 3-8 6-3 7-9 7-1 8-4 6-4 8-2 7-7 4-2 3-0 2-4 2-4 1900 3-2 3-1 5-3 7-2 5-7 7-8 8-0 6-5 5-6 5-1 2-2 2-5 2-2 Whole 1 period .
J 3-0 3-1 4-3 7-2 5-7 6-7 6*4 6-5 5-0 4'2 2*2 2*2 2*2 Corresponding data are not given for V owing to their greater uncertainty .
V ranges are on the average less than D or H ranges , and when very small they are VOL. ccxvi.\#151 ; A. ' Phil. Trans. , ' A , vol. 208 , p. 205 .
2 o exposed to undue uncertainty owing to the large temperature coefficient of the magnet and the absence of a temperature correction .
The uncorrected temperature effect on ft ' is of course not always in one direction , but amongst the least ranges of the month the chances are that some will be reduced by the temperature effect .
Thus the natural consequence of the absence of a temperature correction will be to give too low a value for the monthly minimum range .
The H ranges are not wholly free from this uncertainty , but the temperature coefficient of the H magnet is only about a seventh of that of the Y magnet , while the average range in H is fully 80 per cent , in excess of that in V , thus the uncertainty is of quite a different order .
Table XLYII.\#151 ; Horizontal Force .
Absolute Daily Eanges .
Smallest Yalues ( Unit lie ) .
Year .
January .
February .
March .
April .
May .
June .
July .
' August .
September .
October .
" November .
December .
Year .
1890 16 18 17 31 25 26 31 24 23 25 16 13 13 1891 12 17 18 38 28 35 39 41 40 29 29 13 12 1892 26 36 37 30 41 44 46 45 36 38 19 17 17 1893 20 20 31 52 40 45 42 50 38 36 30 19 19 1894 21 30 25 36 46 60 50 49 31 36 22 19 19 1895 16 29 34 50 45 52 44 27 33 30 21 16 16 1896 24 29 23 43 35 35 .
29 33 48 23 15 12 12 1897 13 15 24 34 36 33 31 28 27 21 18 12 12 1898 10 13 29 16 23 33 28 25 33 23 18 11 10 1899 11 16 22 30 31 27 22 32 37 16 13 9 9 1900 12 12 21 25 26 29 23 22 22 14 11 7 7 Whole 1 period J 10 12 17 16 23 26 22 22 22 14 11 7 7 The D range is entirely free from this source of uncertainty , so far as is known ; it has the further advantage that no error can arise from an unrecognised variation in the scale value .
There is , however , one slight drawback attending it , viz. , that the force required to alter D by V alters if H alters .
Strictly speaking , the force equivalent to a change of 1 ' in D is different at different times of a highly disturbed day .
The slight increase in accuracy that would result from allowing for changes in H in the course of a single day would be no adequate return for the labour involved .
Apart from any disturbance effect , the force equivalent to a change of U in D is influenced by secular change in H. The mean value of H in 1900 exceeded that in FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
1890 by 259y , or about 1'4 per cent , alter D by 1 ' was 5'29y in 5'30y , , 5-31y " 5'32y , , 5'33y , , 5'34y " .
5'35y " 5'36y , , If we go to 3 figures , the force required to 1890 and 1891 1893 and 1894 1897 and 1898 1899 1900 .
This variation , though quite appreciable when we are dealing with large ranges , is hardly worth considering in the case of the minimum ranges in Table XLVI .
Examining that table we see that the winter months , especially December , are conspicuous for small ranges .
The fact that the mean values of B/ for winter months fall conspicuously short of those for other seasons is due , not so much to the absence of large values of R/ , as to the presence of a considerable number of very small values .
In summer , really small ranges are scarce .
In fact , in the half year from April to September no single day of the 11 years had a range under 5 ' .
In 1893 no April day had a range under 12r'5 , though the mean value for the month was only 17'T , and in June the lowest range was 12''3 though the mean for the month was only 16''4 .
The uniformity of the range in June 1893 was very remarkable .
In 21 of the 30 days the range lay between 14/ ,0 and 18/ *5 and the highest value of the month was only 21''6 .
The absolutely smallest range of the 11 years was 2''2 , or 12y in force .
Turning to Table XLVII .
, we see that the lowest range recorded was 7y .
The winter months , as with D , supply all the outstandingly small ranges , December being especially conspicuous .
During the eleven years the sensitiveness of the Kew H magnetograph was always near 1 mm. = 5y .
Thus a range of 7y implies a variation of only 1'4 mm. in the length of the ordinate throughout a portion of curve whose length of abscissa is 360 mm. This means a slope so gentle everywhere that the recognition of the positions of the maximum and minimum is no easy matter .
Again , a change of 1 ' F. in temperature would alter the ordinate fully 0'3 mm. Thus not improbably 7y may have been somewhat an under-estimate of the true minimum range .
At the other end of the scale the largest observed ranges were in D , 1 ' 25/ -6 ( =457y ) on March 15 , 1898 ; in H , 720y on February 13 , 1892 ; in V , 639y on July 20 , 1894 .
It is possible that the D range on March 15 , 1898 , may have been exceeded on 2 o 2 February 14 , 1892 , as the trace then went beyond the limit of registration , and the range measured to the edge of the sheet was 1 ' 19 ' .
It we accept the value on March 15 , 1898 , as the true maximum , we find maximum range/ minimum range = 102'9 in H , as compared with 39'4 in D. S 29 .
Tables XLVIII .
to LIII .
deal with the frequency of occurrence of ranges of different size for D , H , and Y. There are two tables for each element , the one showing the distribution for each of the 11 years , the other the distribution for each of the 12 months .
To make the D results comparable with those for the force Table XLYIII.\#151 ; Declination Ranges ( Unit lie ) .
Number of Occurrences .
Year .
From 0 To 12-5 12-5 25 25 37-5 37-5 50 50 62-5 62-5 75 75 100 100 125 125 150 150 175 175 200 200 250 250 500 500 1890 0 4 55 96 96 68 30 11 4 1 0 0 0 0 1891 0 13 27 35 78 75 80 37 8 5 4 2 1 0 1892 0 1 9 28 49 70 115 39 18 12 9 8 8 0 1893 0 3 13 26 33 61 156 47 15 8 2 0 1 0 1894 0 2 10 36 41 92 107 36 15 4 7 3 12 0 1895 0 6 18 27 49 70 112 41 19 16 7 0 0 0 1896 0 8 26 36 65 82 75 36 18 11 3 5 1 0 1897 1 15 37 57 93 74 45 31 6 2 1 1 2 0 1898 0 17 39 52 102 66 52 22 8 2 2 0 3 0 1899 0 26 38 70 93 64 50 12 3 5 2 2 0 0 1900 1 49 49 96 102 45 14 3 3 3 0 0 0 0 elements , the same force limits are employed for the several classes .
It was not necessary to convert individual D ranges into their equivalents in force , but only to find what were the angles corresponding to the several force limits , employing the relation between 1 ' and lie appropriate to the particular year .
Table XLIX.\#151 ; Declination Ranges ( Unit lie ) .
Number of Occurrences in Eleven Years .
Month .
From 0 To 12-5 1 12*5 25 25 37-5 .
37-5 50 50 62-5 62-5 75 75 100 100 125 125 150 150 175 175 200 200 250 250 500 500 January . . .
0 38 | i 67 74 51 30 37 23 10 6 1 3 1 0 February . .
0 21 !
43 52 44 42 54 26 9 7 4 3 5 0 March . . . .
0 1 11 39 71 68 67 36 24 14 3 1 6 0 April ... .
0 0 0 25 74 85 90 32 12 5 5 1 1 0 May 0 0 6 40 61 82 97 29 14 4 1 5 2 0 June ... .
0 0 2 35 88 89 84 23 2 4 0 3 0 0 July 0 0 2 32 96 97 80 19 3 4 2 3 3 0 August . . .
0 0 1 19 86 93 110 24 4 2 0 0 2 0 September .
0 0 8 36 81 70 87 23 11 3 7 1 3 0 October . . .
0 1 21 69 66 43 68 45 13 8 7 0 0 0 November .
1 30 64 70 50 40 32 16 8 10 6 1 2 0 December . .
1 .53 96 68 33 28 30 19 7 2 1 0 3 0 FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
Table L.\#151 ; Horizontal Force Ranges ( Unit lie ) .
Number of Occurrences .
Year .
From 0 To 12-5 12-5 25 25 37-5 37-5 50 50 62-5 62-5 75 75 100 100 125 125 150 150 175 175 200 200 250 250 500 500 1890 0 40 99 111 66 28 14 5 2 0 0 0 0 0 1891 1 23 49 68 82 52 59 14 11 3 1 1 1 0 1892 0 4 33 66 61 60 68 29 16 4 3 9 10 3 1893 0 10 32 50 81 68 78 25 13 5 0 3 0 0 1894 0 8 44 38 59 71 79 32 12 5 4 2 9 2 1895 0 9 38 57 77 66 68 30 10 8 2 0 0 0 1896 2 26 42 68 72 51 60 26 11 3 3 0 2 0 1897 1 34 69 99 78 39 27 9 6 2 1 0 0 0 1898 3 46 58 80 67 46 41 17 2 1 1 1 0 2 1899 4 45 83 79 69 37 33 9 5 1 0 0 0 0 1900 15 73 117 106 32 11 7 1 1 1 0 1 0 0 Table LI.\#151 ; Horizontal Force Ranges ( Unit lie ) .
Number of Occurrences in Eleven Years .
Month .
From 0 To 12-5 12-5 25 25 37-5 37-5 50 50 62-5 62-5 75 75 100 100 125 125 150 ' 150 175 175 200 200 250 250 500 500 January . .
5 73 97 58 32 32 22 12 ij o 1 1 1 0 February . .
1 45 73 48 49 30 30 14 11 1 2 1 3 2 March . . .
0 12 71 71 68 32 38 21 15 2 o 3 5 1 April ... .
0 2 29 73 64 66 62 19 4 6 2 2 1 0 May ... .
0 1 36 64 80 64 60 18 6 4 2 2 4 0 June ... .
0 0 29 75 78 58 55 14 13 4 1 2 1 0 July ... .
0 2 38 69 78 49 61 24 9 3 1 3 3 1 August . . .
0 5 23 77 94 56 53 21 4 3 1 2 0 2 September .
0 5 32 76 79 52 55 16 7 3 1 0 3 1 October . . .
0 15 61 79 46 48 66 21 1 2 2 0 0 0 November .
3 67 78 78 47 18 17 9 9 3 0 0 1 0 December .
17 91 97 54 29 24 15 8 3 2 0 1 0 0 Table LIT.\#151 ; Vertical Force Ranges ( Unit lie ) .
Number of Occurrences .
Month .
From 0 To 12-5 12-5 25 25 37-5 375 50 50 62-5 62-5 75 75 100 100 .
125 125 150 150 175 175 200 200 250 250 500 500 1890 71 164 111 13 5 1 0 0 0 0 0 0 0 0 1891 26 110 143 39 22 7 7 5 4 1 0 0 1 0 .
1892 13 96 126 58 17 12 15 8 3 3 2 o 9 1 1893 6 106 126 75 23 9 13 3 1 1 1 1 0 0 1894 15 96 134 52 26 12 10 3 2 1 2 3 7 2 1895 13 101 136 53 31 12 13 5 1 0 0 0 0 0 1896 18 127 127 40 21 10 13 7 1 0 1 1 0 0 1897 51 144 119 26 13 2 7 0 2 1 0 0 0 0 1898 37 132 133 40 13 3 3 1 0 0 0 0 2 1 1899 45 170 108 27 5 3 6 0 1 0 0 0 0 0 1900 61 199 94 6 3 0 0 0 0 1 0 0 1 0 Table LIII.\#151 ; Vertical Force Ranges ( Unit lie ) .
Number of Occurrences in Eleven Years .
Month .
From 0 To 12-5 12-5 25 25 37-5 37-5 50 50 62-5 62-5 75 75 100 100 125 125 150 150 175 175 200 200 250 250 500 500 January . . .
76 173 58 14 11 4 4 0 0 1 0 0 0 0 February . .
42 133 74 23 14 5 8 4 0 0 2 1 3 1 March . . . .
5 124 118 37 22 6 14 6 0 3 0 0 5 1 April ... .
1 74 163 51 15 5 14 3 2 1 0 0 1 0 May 0 43 171 67 33 10 5 2 2 2 1 2 3 0 June ... .
0 76 169 48 14 12 5 2 1 0 1 1 1 0 July 0 86 166 48 14 10 8 2 1 0 1 3 1 1 August . . .
2 111 158 46 11 1 5 2 1 1 0 1 1 1 September .
9 139 108 37 14 5 9 3 3 0 0 0 3 0 October .
, .
22 168 91 25 17 8 5 4 1 0 0 0 0 0 November .
74 176 38 19 5 3 7 4 2 0 1 0 1 0 December .
125 142 43 14 9 2 3 0 2 0 0 0 1 0 The step for the six lowest classes is only 12'5y , answering roughly to 2''5 in D , or half the step employed in the corresponding tables in my previous paper .
For the next five classes the step is 25y .
The twelfth class has a step of 50y , the thirteenth of 250y , while the last class includes the few ranges\#151 ; none in D\#151 ; which exceeded 500y .
It requires a large number of classes to show the distribution satisfactorily near the lower end of the scale .
Towards the upper end of the scale , the occurrences are so few that the employment of a very large number of classes with small steps would have been a useless complication .
It would , in fact , require a very much longer period of years to fix the exact law of incidence for ranges exceeding 200y .
To facilitate intercomparison , Table LIV .
includes results for the whole year from the three elements .
The 11 years , the four years representing sunspot maximum , and the three years representing sunspot minimum are treated separately .
Table LIV .
likewise gives results from the three elements for the three seasons , derived from the whole 11 years .
The results in Table LIV .
are shown graphically in fig. 13 .
The number of days in each class was expressed as a percentage of the total number of days included in all the classes , and ordinates were drawn proportional to these percentages , due allowance being made for the difference between the steps in the earlier and later classes .
The graphical representation was not carried beyond the 9th class , whose superior limit is 150y , because the ordinates for the higher classes would have been too short to show satisfactorily , and the irregularities arising from insufficient length of period would have been too great .
In all the curves the range of greatest frequency of occurrence is less than the arithmetic mean range .
Also the range of most frequent occurrence is always greater for D than for H , and much larger for H than for V. On many days when the D and H curves show large irregular Table LIY.\#151 ; Ranges ( Unit lie ) .
Number of Occurrences .
FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
o o ON ^ O LO co O O O O \lt ; M H O CM rH O co cm LO o o 00 CM O rH 05 CO O O rH rH O O O 05 05 N- GO .
CO CM LO CM CM CM \lt ; M rH rH rH i\#151 ; l o o rH I'\#151 ; 00 H rH N ( MHO N- CO rH CO lo C rH 05 N(M \lt ; M CM rH rH rH rH lO o N lO CO lO O lO CM O O cq co co ( MN O CO lo co r-H CM CO rH CM rH CM \#169 ; 10 05 CO 00 O CM LO OMH LO CO rH O CO rH rH CO rH i\#151 ; 1 co co hH CM CM CO rH rH i\#151 ; 1 LO o CM LO n- 05 lo !
\gt ; .
rH N- O 00 rH H O ^ \#169 ; N CO CO DUO rH 00 rH CO LO rH co co co cm CM CO r\#151 ; H r-H 100 125 lO N CM co co 05 CO iO O rH CO 00 CO N- CO LO 00 rH 05 CO CO rH rH CM rH oo r* CO N- rH 05 NCO rH rH j\#151 ; 1 rH lO o n- o r-H CO ^N \#169 ; CO rH rH rH CO CO r-H CM \lt ; M H CM rH 05 CO co co oo 05 05 lo 05 lo LO C0 CM H CM ^ N CM CM 00 LO r^ CM rH CO CM CO ( M LO N- 05 1-H CO LO LO N CO CO O -H CDOOrH rH N\#151 ; CO CO ( MN OICOtH N- N- ^ O rH CO 05 CM CO CM CO CM lO CO 1\#151 ; N- LO CM CM rH rH rH \lt ; M rH CO CM LO H Oi ( MOON rH Nr 00N 05 CM N QO rH O CM .
\#169 ; 'rH N- NNQ OS ZD l\#151 ; l N LO CO C5 LO CO CO CO N \#169 ; CM LO CO 00 N rH rH CM CM rH rH rH CM CM co co LO 05 CM 05 N- rH 00 CM CO CO rHOOO 05 05 O CO LO 05 .
LO CM CM rH rH CO CD O H CO CON CO 05 LO CM 00 O N- O CO LO LO oo r* rH CM CM CM CM CM CM rH CM rH rH CM CM lo i\#151 ; i tH\lt ; N- O N CM CM 05 CO O LO co O CO O i\#151 ; 1 CO .
CM CO LO LO ^ CM tHOh N- hH rH 05 00 rn CM CO LO N- co co co i-H LO rH CM CO CM CO CM rH r^l rH CO CM CO rH LO rH 00 lO CM rH 05 o 00 co CM CO rH ( M hH LO O 00 co .
rH TjH rH CO 05 N LO CO tHI N- CM co o rH CM LO H CO rH co rH LO rH CM CO LO CO rH CM rH lo O CM CM CO CO O O N- rH 05 N- CM CO N OON O O CM T\#151 ; 1 CM LO hH H N- .
CM rH co co rH CO a 8 c PhH a \#169 ; 2 QWf* \#169 ; s V ^ j V ^ i t ^ j t ^ J v ^ j \#169 ; g LO ^ o co 05 co zl m 03 co QQ m \lt ; D s ^ g 8*3 00 rH X \gt ; \gt ; GO \#166 ; .H c3 eg N 05 r^ H\gt ; '"O o ^ .2 ^ \#169 ; c3 2^ ^ \gt ; \gt ; N- 00 LO .2 CM fe ^ *3 CM r^ ^ rH M \lt ; m \lt ; 0 \gt ; H rH cr1^ 2 jg o -5 OS ^ oo , -H o r " ' CO W CO w 2 'S v\#151 ; ' rH V * 05 ^ oo rH Seasons .
Year .
Z5y 3o y 7jy o\amp ; y 73oy 257 507 737 7007 7237 7507 Sunspot mirumuni 77 Years Fig. 13 .
Absolute daily ranges , frequency of occurrence .
FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
oscillations , the V curve shows practically no trace of disturbance , or merely a slight exaggeration of the afternoon maximum .
At times , however , V exhibits disturbances rivalling those in D and H. If we confine ourselves to days when the range exceeded 250y , there were 24 occurrences in Y as compared with 28 in D , and on 4 occasions the V range exceeded 500y , while the largest D range observed was 457y .
During some of these very large disturbances the V trace\#151 ; which has usually a gentfy rounded contour\#151 ; showed large rapid oscillations just like the D and H traces .
The great majority of the largest storms occurred in 1892 or 1894 .
In the group of years of sunspot minimum a range as high as 250y appeared only once in V , and on no single occasion in D or H. There is a certain resemblance between the frequency curves for winter and for the sunspot minimum year , between those for equinox and for the average year , and again between those for summer and for the sunspot maximum year .
The curves for the sunspot maximum year show a less smooth rise to the maximum ordinate than the others .
S30 .
Fig. 14 presents the range distribution data from a different point of view .
Unit abscissa represents the arithmetic mean range for the element and season considered .
Thus , in the case of the year from the 11 years , it represents 72'2y in D , 60'7y in H , and 32'9y in Y. This eliminates the effect of mere size , and so helps to bring out the degree of resemblance between the laws of distribution followed in different elements , or in the same element at different seasons .
In the case of the whole year , the range of greatest frequency occurs very nearly at the same place\#151 ; i.e. , bears very nearly the same ratio to the mean range\#151 ; in the H and Y curves ; in D it answers to a decidedly larger abscissa .
In the case of H , the range of greatest frequency answers to a higher and higher fraction of the mean range as we pass from winter to equinox , and from equinox to summer .
There is not much difference between the curves in fig. 14 answering to sunspot maximum and minimum , though the mean ranges are respectively 75'8y and 43'9y .
All the frequency curves have the general aspect of the ordinary diagram of intensity of visible and invisible radiation from a solid at high temperature .
This , however , may mean no more than that the number of classes in which the ranges were grouped was too limited to show the existence of bands of reduced frequency or of no frequency .
The existence of such bands , at least towards the side of the very high ranges , is by no means improbable , especially if storms should prove to have more than one origin .
For the careful measurement of the curves and the calculation of the hourly means I am indebted to various members , past and present , of the Kew Observatory staff .
Amongst these I would particularly mention Messrs. G. Badderly and C. Cooper , vol. ccxvi.\#151 ; a. .
2 p Meaji .
Whole from 11 years ft d Fig. 14 .
Absolute daily ranges frequency of occurrence .
FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900 .
now of the National Physical Laboratory , and Mr. B. Francis , the present magnetic observer at Kew Observatory .
During the eleven years whose records are mainly considered the magnetographs were under the charge of Mr. T. W. Baker , then Chief Assistant , who also took the great majority of the absolute observations .
The * homogeneousness of the material owes much to Mr. Baker .
The expense of measuring the curves was defrayed mainly out of grants from the Government Grant Committee .
The calculation of the diurnal inequalities , Fourier coefficients , and other arithmetical work has been mainly done in my leisure hours .

