]\gt ; IV .
The ) iurnal Tariation of fognetism .
By ARTHUR SCH tSTER , F.R.S. Received October November 7 , 1907 .
1 .
IN a previous communicatio I proyed that the Diurnal Variation of Terrestrial Magnetism has its origin outside the earth 's surface and ) natural conclusion that it was caused by electric currents ting in the upper regions of the atmosphere .
If we endeavour to carry the investigation a step further and enquire into the probable origin of these currents , we } ) at prese1lt no altel.native to the theory first proposed by BALFOUR that the necessary electromotive forces are supplied by the permanent forces of terrestrial magnetism acting ] bodily motion of masses of conducting air which cut through its lines of force .
In the language of modern electrodynamics the periodic netic disturbance is due to Foucault currents induced in an oscillating atmosphere by the vertical Iletic force .
The problem to be solved in the first instance is the specification of the internal motion of a shell of air , which shall , under the action of given magnetic forces , determine the electric currents producing known electromagnetic effects .
Treating the diurnal nd semidiurnal variations separately , the calculation leads to the interesting results that each of them is caused by an oscillation of the atmosphere which is of the same nature as that which causes the diurnal changes of barometric pressure .
The phases of the barometric and magnetic oscillations agree to about hours , and it is doubtful whether this difference may not be due to uncertainties in the experimental data .
In the vious communication referred to I already tentatively suggested a connexion between the barometric and magnetic changes , but it is only recently that I have examined the matter more closely .
In the investigation which follows I begin by considering the possibility that both variations are due to one and the same general oscillation of the atmosphel.e .
The problem is then absolutely determined if the barometric ange is known , and we may calculate within certain limits the conducting power of the air which is sufficient and necessary to produce the rved magnetic effects ; this conducting power is found to be considerable .
It is to be observed , however , that the electric currents producing the magnetic variations circulate only in the upper layers of atmosphere , where the pressure is too small to affect the barometer ; the two ' Phil. Trans , vol. 180 , p. 467 ( 1889 ) .
430 .
22.4.08 MR. ARTHUR SCHUSTER ON THE variations have their origin therefore in different layers , which may to some extent oscillate independently .
Though we shall find that the facts may be reconciled with the simpler supposition of one united oscillation of the whole shell of air , there are certain difficulties which are most easily explained by assuming possible differences in phase and amplitude between the upper and lower layers .
If the two oscillations are quite independent , the conducting power depending on the now unknown amplitude of the periodic motion cannot be calculated , but must still be large , unless the amplitude reaches a higher order of magnitude than we have any reason to assume .
The mathematical analysis is simple so long as we take the electric conductivity of the air to be uniform and constant ; but the great ionisation which the theory demands requires some explanation , and solar radiation suggests itself as a possible cause .
Hence we might expect an increased conducting power in summer and in daytime as compared with that found during winter and at night .
Observation shows , indeed , that the amplitude of the etic variation is considerably greater in summer than in winter , and we know that the needle is at comparative rest during the night .
The variable conducting power depending on the position of the sun helps us also to overcome a difficulty which at first sight would appear to exclude the possibility of any close connexion between the barometric and magnetic variations ; the difficulty is presented by the fact that the change in atmospheric pressure is mainly semidiurnal , while the greater portion of the magnetic change is diurnal .
This may to some extent be explained by the mathematical calculation , which shows that the flow of air giving a 24-hourly variation of barometric pressure is more effective in causing a magnetic variation than the corresponding 12-hourly variation , but the whole difference cannot be accounted for in this manner .
If , however , the conductivity of air is greater during the day than during the night , it may be proved that the 12-hourly variation of the barometer produces an appreciable periodicity of 24 hours in the netic change , while there is no sensible increase in the 12-hourly magnetic change due to the 24-hourly period of the barometer .
The complete solution of the mathematical problem for the case of a conducting power proportional to the cosine of the angle of incidence of the sun 's rays is given in Part II .
But even this extension of the theory is insufficient to explain entirely the observed increased amplitude of the magnetic variation during summer .
We are , therefore , driven to assume either that the oscillation of the upper layer is greater in summer than in winter and is to that extent independent of the oscillation of the lower layers , or that the ionising power of solar radiation is in some degree accumulative and that atmospheric conductivity is therefore not completely determined by the position of the sun at the time .
The increased amplitude at times when sunspots are frequent is explained by an increased conductivity corresponding to an increase in solar activity .
All indications , therefore , point to the sun as the source of ionisation , and ultra-violet radiation seems to be the most plausible cause .
DIURNAL VARIATION OF TERRESTRIAL I. 2 .
The velocity potential of a ) horizontal irrotational motion of the earth 's atmosphere considered as an infinitely thin shell is necessarily expressible as a series of harmonics from which we may select for consideration the one of degree and type , writing it or according to convenience .
The longitude is measured from any selected meridian towards the east , while is the time of the standard meridian in measure .
As in the greater part of the investigation and occur in the combination only , we may frequently omit without detriment to the clearness , noting , however , that if the differentiation with respect to is replaced by a differentiation with respect to we must apply the factor , where is the number of seconds in the day .
I consider in the first instance the electric CUl.rents which are induced in air moving horizontally under the influence of the earth 's vertical magnetic force .
Assuming the earth to be a uniformly lnagnetised sphere , its potential may be resolved into the zonal harmonic of the first and the tesseral harmonic of the first type and degree .
The angle between the lnagnetic axis and the geographical axis not being great , the zonal harmonic constitutes by far the largest part , and forms the first subject of our investigation .
As .
as this part is concerned , we may put the vertical force equal to , where is the colatitude and , measured upwards , has a numerical value differing little from The components of electric force , X and , measured towards the south and east respectively , are . . . . .
( 1 ) , and these ions may be written in the form ] .
( 2 ) , where is the degree of the harmonic .
The identity between ( 1 ) and ( 2 ) is obvious as ards the first of the equations , and the reduction of the second is obtained with the help of the fundamental equation being a zonal harmonic of degree The components of electric force are by ( 2 ) reduced to the fornl ; ( 3 ) , and may be divided into two portions , the first of which is deriyable from a potential .
In the steady state this part is balanced by a static distribution of electricity revolving round the earth and using a variation in the electrostatic potential which is found to be too weak to affect our instruments .
The second portion the electric force produces electric currents ; these , electric inertia\mdash ; which will be considered later\mdash ; have as current function , where is the conductivity of the medium .
The comparison of ( 2 ) and ( 3 ) shows that , and by means of well known reductions may be expressed in the normal form . . .
( 4 ) .
Here and are the two spherical harmonics of degree and type which have the same numerical factor as the current function I shall confine myself to the two principal portions of the diurnal variation of pressure which associated with the velocity tials and The corresponding electric current functions seen by ( 4 ) to be and . . . . .
( 5 ) .
It is shown by CLERK MAXWELL Electricity Magnetism , that the magnetic forces accompanying the currents in spherical sheets which are derivable from a current function having the of a surface harmonic are obtained from a magnetic potential which is equal to the same harmonic multiplied by a factor which inside the spherical shell is .
The thickness of the atmosphere being negligible compared with the radius of the earth , we may put , and obtain thus , for the magnetic potential due to the induced electric currents , sin ] ( 6 ) .
represents the thickness of the shell of the conducting layer , and is introduced because the current functions used above yield current densities , while MAXWELL'S result applies to functions which lead directly to currents .
Our represents the potential of the diurnal variation of terrestrial magnetism calculated from an atmospheric oscillation according to our theory , and agrees in form OF TISAI .
with the principal terms of that variation as observed when the average annual value is considered seasonal changes are disregarded .
3 .
It remains to be seen whether the calculated variation agrees as regards phase and can be made to coincide in magnitude by a reasonable value of the conductivity and thickness of the effective layers of the atmosphere .
For this purpose we first obtain a value for the constants and .
If Sp be the variation of the pressure , and the corresponding change of the density , we have where is the velocity potential and the velocity of sound .
Under the assumption that the whole atmosphere oscillates equally in all its layers , Sp/ p will be the same at every point of a vertical line , and we may , therefore , determine its value at the surface of the earth .
According to HANN ( ' Meteorologie , ' p. 189 ) , the change of the barometer at the equator , measured in millimetres , is represented by If this expression be denoted by , we must assig1l the value of 760 to to brin the units into nony .
It follows that at the equator . . .
( 7 ) .
The numerical value of is , or , where is the radius of the earth .
The constants and of the velocity potential in ( 6 ) are , therefore , determined by , and ( 8 ) .
We ultimately get for the calculated magnetic potential cos2 or , introducing the value of and restoring the term containing the latitude , ( 9 ) The principal terms of the diurnal and semidiurnal variations of magnetic force , abstraction being made of seasonal changes , were found in my previous communication to be ( 10 ) .
If we compare the phases , we find that the magnetic potential calculated from the barometric variation lags behind the observed one by an amount which is 1 hour and 36 minutes for the diurnal and 1 hour and 50 minutes for the semidiurnal variation , showing a remarkably close agreement in the two terms .
As regards amplitude , we can establish agreement by adjustin the value of , but the same value ought to satisfy both terms , which is not the case , being if calculated from the 12-hourly variation , and if calculated from the 24-hourly variation .
4 .
The tacit assumption has been made that the barometric variation is distributed over the surface of the earth according to the simplest harmonic term consistent with each period , so that for the diurnal variation the amplitude would be proportional to the cosine of the latitude , and for the semidiurnal variation to the square of the cosine .
The experimental data show , however , that there are other terms to be considered , and for the semidiurnal variation Dr. ADOLF SCHMIDT has obtained the best agreement by introducing the harmonic of the fourth having an amplitude which at the equator would amount to the twelfth part of the whole effect .
The amplitude of the second term in Equation 9 would consequently be reduced to , and a higher harmonic would be introduced ; but the experimental evaluation of these higher terms in the magnetic variation is too uncertain to be taken account of at present , their effect in any case bein small .
As regards the hourly variation , its dependence on latitude has not been clearly established .
The term , as commonly observed , is much affected by local circumstances , and HANN takes therefore the observation on board ship to represent the true phenomenon so far as it depends on the atmosphere as a whole .
While greater consistency is thus gained , the observations on board ship cannot lay claim to the same accuracy as those taken on land , and , as the figures show , considerable uncertainty still prevails .
In the following table the values given by HANN are collected together:\mdash ; These numbers do not follow any very simple law and can only be very partially represented by an expression varying as the cosine of the latitude .
The rapid increase in the amplitude at latitudes of about suggests the presence of the third harmonic , and treating the figures by the method of least squares we are led to an expression If this equation were to represent correctly the distribution in latitude of the DIURNAL VARIATION OF TERRESTRIAL MAGNETISM .
169 diurnal term , the calculated amplitude of the magnetic potential would be increased considerably more than is required to bring it into harmony with the semidiurnal term , because not only is the amplitude of the term of the barometric variation in increased , but the additional term gives rise to a magnetic potential which is of the same degree and type and more than doubles the effect of the first term .
Very little importance , however , can be attached to this calculation , which depends to a great extent on the last entry of the foregoing table ; but enough has been said to show that our present knowledge of the 24-hourly variation of the barometric pressure is very uncertain , and that a term of the third in its expression is likely to diminish materially the discrepancy between the electric conductivity of the atmosphere as derived from the diul.nal and semidiurnal periods .
5 .
We must next turn our attention to seyeral corrections which modify the calculated values without , however , introducing material changes .
The observed magnetic variations have been treated as if they were wholly due to outside causes , although it was shown in my previous communication that an appreciable portion of it was an effect of electric currents induced inside the earth by the varyin potential itself .
What we observe is the resultant of the original outside effect and its concomitant induced inside effect .
To explain the absence of time lag of the induced variation it was necessary to assume a good conductivity of an inner core and small uctivity of the outer shell .
An estimate may be made of the radius of the conducting core .
If the outer potential is represented by , where is the radius of the earth , and the inner potential is , an estimate of may be obtained by the fact that if the inner conductivity is sufficiently great the vertical force is entirely destroyed at the surface of the inner core .
If this has a radius it follows that and must have equal values , or that The resultant potential at the surface of the earth is , therefore , If this expression is multiplied by , we obtain the vertical force calculated from the horizontal force on the supposition that the whole effect comes from the outside .
The observed vertical force , on the other hand , is The previous result showed that the actual vertical force was about half the calculated one , the principal term being that due to .
We find in this way The thickness of the outer non-conducting crust would thus appear to be about 1000 kilometres , and cannot , therefore , be connected with the layer having a thickness VOL. CCVIII .
of about 30 miles which shows itself in its effect on seismic waves , and , according to STRUTT , contains the radioactive matter .
On the other hand , it is quite likely that the outer sbell is identical with that which the discussion of the propagation of seismic waves shows to have different elastic properties from the nucleus , and which , according to recent rehes , has a thickness of 1500 kilometres .
The observations show that the internal potential has a value equal to one-fourth of the external one , or that the external potential represents of the whole .
For the same value of we similarly find that the outside effect is times that of the .
The coefficients in equationl ( 9 ) Should therefore be diminished by multiplying with and 0 respectively .
6 .
We may now complete ] investigation as far as it relates to uniform conductivities .
The magIletic and )gral ) hical poles of the earth not coinciding , the vertical force is not simply proportional to the cosine of the colatitude , but a term must be added to , where is measured from the meridian ' west of Greenwich , which is that containing the magnetic axis .
I discuss the effect of the inclination of the magnetic axis somewhat in detail , as it will give us a good test of the proposed theory when suitable observations will be available .
eaving out the factor , where represents the vertical force at the pole and the colatitude of the magnetic pole , the electric forces , as far as they concern us at present , are ; If these values be substituted in ( 3 ) , the elimination of leads to the equation .
( 11 ) .
It will be shown in Part II .
that is equal to where ; ; ; . .
The only case necessary to consider is that in which , when , so that we may disregard the factor .
If , then ; and if , then ; % ; DIURNAL VARIATION OF TERRESTRIAL TISM .
We can satisfy equation ( 3 ) by assuming to be made up of two or four terms , according as we treat of the diurnal or semidiurnal variations .
Remembering that by the fundamental equation we find the terms which are introduced by the inclination of the magnetic axis to be the diurnal variation , and for the semidiurnal variation .
The additional terms are therefore , restoring the constant factor , . . . . .
( 12 ) .
Here represents the angle between the magnetic and geographical axes , and has been introduced to represent the local time functions are the tesseral functions , so that Equations ( 12 ) show that if the inclination of the magnetic axis be taken into account , the diurnal variations do not entirely depend on local time .
A barometric variation of a certain period is accompanied by a magnetic variation of the same period , as is obvious ; but if the lines of equal magnetic potential in the diurnal variation are drawn as in my previous communication , a barometric variation represented by results in a magnetic potential containing terms These equipotential lines and their coincident stream lines revolve with velocities and round the earth , being its angular velocity , and in this way variations proportional to are produced .
In order to estimate the magnitude of these terms , consider the diurnal variation , the normal term of which has been found to be equal to .
Along a meridian circle for which is either or , the additional terms are , putting equal to unity , for and for .
In these equations the numerical value has been introduced for At the equator the additional terms , therefore , have amplitudes and respectively , as compared with the main diurnal term .
The force to geographical west being proportional to the potential , we may take these numbers to be the amplitude of the westerly force variations .
The main variation is proportional to , and has zero value at the equator in the tropical region .
The additional terms are therefore the ruling terms at the equator .
The horizontal force along the same circle has unit amplitude , measured on the same scale , so that the new terms come well within the range of our observational powers .
It would be interesting to trace them , but it should be remarked that only observations made near the equinox are suitable for the purpose , as the seasonal terms , which yet remain to be discussed , would otherwise interfere .
7 .
We may interrupt the progress of our investigation for a moment to inquire into the magnitude of the electrostatic effect dependent on the potential which was found equal to , leading to a vertical electric force .
In the cases which concern us , , and has values at the equator which were found to be and 15 respectively .
It follows that the variation of vertical electric force is of the order of 3 C.G.S. units , which is only 1 volt per 300 kilometres .
This may be disregarded .
8 .
The previous discussion has only taken the earth 's vertical magnetic force into consideration .
The horizontal force causes , in combination with a horizontal atmospheric oscillation , a vertical electromotive force , and so far as this produces electric currents , their flow is in opposite directions in strata which are vertically above each other .
The magnetic effect is therefore of a smaller order of magnitude than that due to vertical force .
9 .
In calculating the currents from the electric forces , I have applied OHM'S law , and therefore neglected the effects of electric inertia ; but it is not difficult to estimate the change of phase which results from self-induction .
Using the equations given by for spherical current sheets , we find that if is the function defined by equation ( 3 ) , and the current function , where is the conductivity , and ; provided that is a surface harmonic of degree .
If the latter function is proportional to , we find in the usual way * See CLERK MAXWELL , ' Electricity and Magnetism , ' vol. II .
, S672 .
DIURNAL VARIATION OF TERRESTRIAL MAGNETISM .
If we take the current sheets to be of finite thickness , and denotes conductivity referred to unit volume , we must write If has the value previously determined by the semidiurnal variation , I calculate a retardation of 1 hour for the semidiurnal and about hours for the diurnal term .
The amplitude would be reduced by about 14 per cent. There are various causes , notably the inequalities of conductivity , tending to diminish the retardation , so that we may consider that self-induction would not cause a shift of phase amounting to more than an hour , but it is in the opposite direction to that indicated by the observations , if the barometric and magnetic oscillations are due to identical causes .
10 .
It is known that the air contains an excess of positive electricity , and the question might be raised whether the oscillations of the atmosphere do not convectively produce direct magnetic effects .
If be the total quantity of electricity contained in a vertical column of unit cross section , and , the velocity of air in that column , supposed to be uniform , the total current in the atmosphere across a vertical area of unit width is , and the magnetic force at the surface of the earth is of the order of magnitude .
The quantity must be equal and opposite to the surface electrification of the earth , which itself is equal to , where is the velocity of light , and the normal fall of potential , which we may put equal to 1 volt per centimetre , or to .
The magnetic force has therefore a magnitude of order .
If the velocity potential of the atmospheric oscillation is , the velocity in the two cases considered is greatest at the equator , where its maximum rises to , which for the diurnal and semidiurnal change is and respectively .
It follows from the numerical values given in ( 8 ) that the maximum equatorial velocities are 10 and 30 centims. per second respectively .
The magnetic forces due to such velocities are quite insignificant .
In the literature referring to the subject we frequently find it suggested that magnetic disturbances are due to moving masses of electrified air , some writers even going so far as to say that this has been proved ; it may be demonstrated , on the contrary , that the assumed cause is insufficient .
For horizontal air currents this has just been demonstrated , and the effects of ascending or descending currents are still less efficient .
If a column of air of cylindrical shape having as base a circle of radius rises or falls with velocity V , and it is imagined as an extreme case that the column extends indefinitely in both directions , the magnetic force at the boundary is , where is the electric density .
At the surface of the earth the ionisation is such that the free electric charges of each kind amount to about 1 electrostatic unit per cubic metre .
Let us assume one kind to be suppressed altogether , so that this number represents the electric volume density , or in electromagnetic measure .
If be 1 kilometre , and the velocity that of an express train , or 30 metres a second , the magnetic effect would be .
This is insignificant and leaves a good margin for a greater sectional area of the ascending current , especially if it is remembered that both our assumed velocity and the volume charges are many times greater than is allowable .
Magnetic due to the motion of electrified air must therefore be ruled out as effective causes of either regular or irregular magnetic changes .
11 .
The daily variation of the magnetic forces includes a strong seasonal term , the amplitudes being greater in summer than in winter .
In order to explain this term according to the theory advocated , it is necessary to assume a greater electric conductivity of the atmosphere in summer than in winter , or an oscillation of greater amplitude , which is not , however , indicated by the barometric changes .
That the conductivity depends on the position of the sun , and may therefore vary with the season , is suggested by the relation in phase between the diurnal and semidiurnal terms , these terms combining together so as to leave the needle comparatively quiescent during the night .
Reserving the possible causes of the conductivity and its dependence on solar position for further discussion , we may complete the theoretical investigation by introducing a variable conductivity .
The simplest supposition to make will be that the conducting power in any small volume is proportional to the cosine of the angle between the vertical and the line drawn to the sun , or , in other words , proportional to the the angle , measured at the centre of the earth on the celestial sphere , between the sun and the small volume considered .
This angle is expressed in spherical co-ordinates by S S . . . . .
( 18 ) , where is the longitude measured from the meridian passing through the sun , and is measured from the pole , S representing the sun 's declination .
To put the assumed law of conductivity into mathematical form , we write If , the conductivity would be zero at a poin opposed to the sun , and this is the highest admissible value of .
In order to keep our investigation as general as possible , I write , where and may have any assigned values .
The solution of our problem is obtained if we can find values of and satisfying the equations ( 14 ) .
will now give directly the current function which hitherto was denoted by DIURNAL yARIATION OF TERRESTRIAL MAGNETISM .
The general problem will be treated in the Appendix , where it is shown that for practical purposes and may be treated as small quantities , the squares of which may be neglected .
The equations may then be written sin ( 15 ) .
Neglecting and , our previous results give in terms of .
Let be one part of thus obtained .
The next approximation is found by substituting for in the terms of ( 15 ) , which contain and The complete value of , as far as it depends on , will then be , where ' is determined by ( 16 ) .
In the two cases which specially interest us we must substitute for the values respectively of and as determined by ( 5 ) .
The solution of ( 16 ) involves the elimination of Treating the terms containing and separately , we find for as far as it depends on if ' is expressed as a series of harmonics , being the degree of one of the terms of the series .
The left-hand side may be transformed as shown in the Appendix , the result being given by ( 25 ) ; we obtain in this way is therefore expressible by two terms , and , so that As regards the terms in , the elimination of leads to The left-hand side is transformed , as shown in the Appendix , the result being given in ( 22 ) .
We thus find expressible as a sum of four terms\mdash ; For we must substitute and when treating of the diurnal and semidiurnal variations respectively , where and are the harmonics of type and degree indicated which have the same factors as the current functions and To the magnetic potential , a further multiplication by is required .
We see that each barometric variation now leads to six terms in the magnetic potential , the factors of being collected in the following table : \mdash ; DIURNAL VARIATION .
Velocity Potential : .
Magnetic Potential : Values of : SEMIDIURNAL VARIATION .
Velocity Potential : .
Barometric Variation : Values of Vertical force at geographical North Pole , measured upwards .
Electric conductivity of atmospheric shell .
Thickness of atmospheric shell .
IUliNAL yARIATION OF TERRFSTRIAL TIShI .
The main terms of the magnetic potentials and are now each affected by both the diurnal and semidiurnal barometric variation , and their relative amplitudes may differ considerably from those calculated on the assumption of a uniform conductivity .
If has its maximum value , which is unity , we have for these two terms , an unimportant difference of phase , and leaving out common factors , 4 .
The 24-hourly variation of magnetism ow takes the lead and as regards westerly force is now times as as the semidiurnal variation , but the latter is still too great for complete agreement with the facts , the observational ratio This remaining discrepancy is not decisive against the accuracy of the assigned cause in view of the uncertainty which attaches to the 24-hourly term in the barometric variation as explained in S 4 , and the considerations brought forward in ] following paragraph .
There is the theoretical possibility of a further increase in the diurnal term through a velocity proportional to ; the motion specified by this potential would give a barometric oscillation determined by the time of some definite meridian , and though observations seem to indicate the existence of part of the oscillation being of this nature , it is not likely that it is sufficiently great to produce a marked magnetic effect .
12 .
A general review of the argument , even at the risk of repeating a portion of what has already been said , may be appropriate , and is necessary to show how we are naturally led to the theory here posed .
It will also serve to oduce the consideration of the remaining difficulties and of the possibility of accounting for the amount of ionisation necessary to ) lain the magnitude of the observed effects .
Our object is to explain the cause of the periodic changes of the terl.estrial magnetic forces in so far as they depend on the position of the sun .
diurnal changes may be represented as being governed by a magnetic potential composed of terms of the form , where is a surface harmonic ; the obser vertical forces show , as proved in my previous communication , that we must seek the cause of the variation outside the earth 's surface .
Electric rents circulating in our atmosphere and having a current function made up of terms which are respectively proportional to produce the required effect , and we are justified in assuming this\mdash ; the simplest explanation\mdash ; to be also the correct one until it is shown to lead to contradictions .
The maintenance of the electric currents necessarily requires an electromotive and their closed lines of flow dispose.of any theory which would seek this force in a static distribution of potential .
charges along by air currents have VOL. CCVIII .
, 2 been shown to be insufficient to produce appreciable effects , and we are therefore driven to look upon electromagnetic induction as being the only possible cause of the observed effects , the earth 's magnetism and atmospheric circulation being the active agents .
Assuming as most probable that atmospheric circulation is symmetrical north and south of the equator , the character of the netic variation shows that the effective component of terrestrial magnetism has opposite signs in the two hemispheres ; it must therefore be the vertical component which is active .
We next put the question : Wbat must be the atmospheric circuIation which under the action of the vertical magnetic force produces periodical magnetic effects equal to those actually observed ?
Taking the average for the complete year , the leading terms of the variable magnetic potential are and , the amplitude of the diurnal term being equal to about eight times the amplitude of the semidiurnal one .
Calculation shows that may be produced by a quasi-tidal atmospheric flow having as velocity potential either or , while the semidiurnal term may produced by a flow having a velocity potential or .
But these velocity potentials are exactly what is required for the atmospheric waves causing the daily changes of barometric pressure .
The semidiurnal term of the pressure change is the one least affected by local conditions , and its distribution over the earth 's surface is therefore accurately known .
It is found that is small compared with .
As regards the diurnal term having an amplitude at the equator of only one-third of the semidiurnal one , it varies somewhat ularly and the relative importance of and is not well ascertained .
Assuming the barometric variation to be wholly due to and , we may deduce the magnetic variation and compare it with the observed changes .
This has been the course of the investigation in the preceding paragraphs .
It is found that the calculated magnetic variations have a phase which lags behind the observed one by about hours , and this lag is slightly less for the diurnal term , but the difference is insignificant in view of the uncertainties of the data .
The amplitude of the calculated diurn al term is about times as great as that of the semidiurnal one , while observation gives , as has already been stated , a ratio of 8 for the two terms .
But if part of the barometric variation is due to a term \mdash ; and there is some evidence that this is the case\mdash ; agreement in the ratio of the two terms may be secured .
There is , however , a further cause tending to increase the semidiurnal magnetic variation .
In order to explain , on the basis of our theory , the difference in the magnetic changes between summer and winter , we must assume that the conductivity of the atmosphere is greater in that hemisphere which is more directly under the influence of the solar rays .
Assuming that the electric conductivity is proportional to , where is the angle measured on the celestial sphere between the sun and the point considered , the calculated semidiurnal term reaches a value which is times as great as that of the diurnal term , so that the term is now called upon to a much smaller extent for making up the deficiency in the diurnal term .
The posed inequalities of DIURNAL yARIATION OIi RRESTRIAL MTISM .
conductivity , though a better agreement between } diurnal and semidiurnal terms , are insufficient to account completely for the large excess of the summer variation over that observed in winter .
This inequality is expressed by for the semidiurn al variation and for the diurnal variation , and its relative magnitude is indicated by the atios and respectively .
The calculated value of both ratios is shown by the tables in S 10 to be S , where S is the sun 's declination .
If we compare the variations during the six summer months with ] those .
the six winter , we must substitute for S its avel.age value , which is about .
On the other hand , the results of my previous communication allow us to deduce the ratios and from the observations , and we find in this and respectively , or values between two and three times as great as those calcuIated from the assumed law of conductivity .
To explain the difference we might imagine some cumulative effect , so that in midsummer the conduction would be greater than in winter even for the same eleyation of the sun , but present knowledge does not justify us in assuming this to be the case .
I am inclined , therefore , to consider that the cause of the discrepancy lies in the fact that , as already suggested in the introductory remarks , the oscillations responsible for the barometric and magnetic phenomena are to some extent independent of each other , affecting different layers of the atmosphere .
There are theoretical reasons why this should be so .
It is now , I think , generally nised that the importance of the selnidiurnal variation of the barometer is due to the fact that the free period of the atmospheric oscillation , dependent on the velocity potential , is very nearly equal to 12 hours .
But it is to be remarked that if concentric layers of the atmosphere be considered separately , there must be a considerable variation in the free periods owing to differences of temperature , and in the highest regions , in which alone electric currents of sufficient intensity can circulate , the temperature is probably so low that the free periods would be more than doubled .
If we take these highest layers to oscillate to some extent independently , we should not therefore find the semidiurnal variation stand out in the same way as it does for the lower layers .
Further , the inequalities of solar radiation in the two hemispheres near solstice ought to cause an appreciable oscillation dependent on the velocity potentials and The barometric variation due to is unimportant compared with that due to because although the reed period is 12 hours , the free period corresponding to the motion inyolyed in it has now a different value ; } ) in the upper layers the relative importance of would be increased , or , as it would be more correct to say , the relative importance of disappears .
This would account for the magnitude of the seasonal term in the magnetic variation .
The suggested partial independence of the oscillations of the upper and lower atmospheres may also explain the discrepancy of phase , which we found to be hours , but is in reality somewhat greater , owing to the fact that self-induction has been neglected in calculating the phase .
With the calculated conductivity , self-induction would cause a retardation of about one hour if the amplitude of oscillation is that deduced from the barometric .
If the amplitude in the regions is reater , the effect of self-induction is correspondingly less , because a smaller conductivity would then be required to account for the magnetic change .
A few words should be said on the uncertainties of the data which serve as a test of the proposed theory and which are derived from my previous calculation of the variation potential .
In deducing that potential I was practically obliged to confine myself to the records of four observatories ( Bombay , Lisbon , Greenwich , St. Petersburg ) , all four being situated in the northern hemisphere ; and the year 1870 being the only one for which records were available at all four stations , I had to base my calculations on the figures for year .
Unfortunately , 1870 was a year of unusual sunspot activity and the magnetic records for that year cannot taken as quite normal .
It is probable that if the average of a nulnber of years were taken , the phases of the components and their relative amplitudes might be somewhat altered ; but I do not think that , as far as the averages for the whole year are concerned , the results of the present investigation wouId be materially altered .
A renewed discussion is , however , very desirable , especially if ) in the southern hemisphere could be made use of .
In my preyious calculations I separated the summer from the winter months , and assuming what is known to be approximately the case , each hemisphere to behave alike when the sun occupies corresponding positions , I was enabled to form an expression of the potential applicable for the whole world simultaneously .
But this is admittedly a defective process , and in drawing the equipotential curves I was careful for this reason to use only the averages taken over the whole year and to make no attempt to separate the two hemispheres .
BEZOLD , who later , basing his calculations on my figures , effected this separation , is often quoted as having thus completed my own investigation , but his extension of my work , for the reason given , seems to me to be to push too far the observed approximate symmetry in the two hemispheres .
What I now regret , however , is that I did not divide the ' into four parts instead of two , as Dr. CHREE'S results seem to show that the time of equinoxes special consideration .
If the views here brought forward are correct , all peculiarities of the barometric variation should be reproduced in the netic effect , though we must remember that the converse does not hold , and that peculiarities of the magnetic effect depending partly on variations of electric conductivity need have no counterpart in the barometric changes .
Thus the greater amplitude of the magnetic variation between summer and winter has already been ascribed to increased conductivity of the atmosphere during the summer .
The .
close relationship between the two phenomena is confirmed by the increased amplitude observed in both near the time of equinoxes .
The diurnal period of barometric pressure is known to have maxima at these epochs , and the valuable researches of Dr. have shown that these maxima are also found in the diurnal variation of the magnetic element .
If we take DIURNAL yARIATION OF TERRESTRIAL MAGNETISM .
the variation of declination as characteristic , .
CHREE'S formula for the semidiurnal term , leaving out the annual variation , is : S where is measured from beginning of January and each month counts as 30 rees .
The corresponding term in the barometric formula is , according to HANN , but if I understand the formula correctly , the time here is counted from the middle of January .
To make the equations correspond , we must therefore diminish the angle in the last equation by 30 degrees , reducing it to degrees , in close agreement with that given by CHREE , the phase angle for the equinox being 265 rees .
The maxima of horizontal force occur , however , a fortnight later , so that too much value should not be given to this agreement ; the effect in amplitude is only about half for the barometric variation ; but questions of conductivity may affect this ratio .
A remarkable feature distinguishing the barometric change is the maximum which takes place simultaneously in both hemispheres early in January when the earth is in perihelion .
According to the theory here discussed , a corresponding annual inequality should show itself in the magnetic variation , though the effect would be partially masked in the northern hemisphere by the changes of conductivity , and could only be ascertained by a comparison of the annual terms in the two hemispheres .
We should expect the difference between winter and summer to be more marked in the southern hemisphere , because there the effects of conductivity would act in the same direction as the effects of diminished distance from the sun .
It is much to be desired that some systematic attempt should be made to investigate the lunar influence on the netic changes , for we possess at present only the vaguest information as to how the diflerent components of magnetic force are affected .
It is quite possible that the effects may depend on a tidal disturbance of the upper regions of the atmosphere .
If so , we may expect to get a valuable test of our theory by their investigation .
13 .
We are now prepared to discuss the magnitude of the conductivity required in order that the proposed theory should be tenable .
If equations ( 9 ) and ( 10 ) are compared with each other , and the correction discussed in 5 be applied , we find from the semi-diurnal term The first question which arises is the value to be assigned to .
Obseryations of the aurora borealis conducted by the Danish expedition under the late Mr. PAULSLN have led to the conclusion that the arc of these luminosities is generally at a height .
Zeitschrift , ' 1898 , vol. , p. 381 .
Mh ' .
ARTHUR SCHUSTER ON THE of 400 kilometres .
* The height of meteors when they become luminous is as a rule over 100 kilometres , } ) there is one case on record in which a height of 780 kilometres was found .
We may therefore take 300 kilometres as an outside limit for , giving the value of as the lower limit for the conductivity .
This no doubt is a high value , and there may be some hesitation in accepting it as a possible one .
Mr. C. T. R. WILSON has , however , ] ready drawn attention to the fact that at high altitudes we must , with the same ionising power , expect a much increased conductivity , for the ionic velocity due to unit difference of potential varies inversely with the pressure .
If , further , as the experiments indicate , the ionising power and rate of recombination of ions both diminish directly as the pressure , it would follow that when the pressure is only the millionth part of an atmosphere the conductivity should be for the same ionisation a million times as great as at the surface of the earth .
Researches on the conductivity of gases generally give relative measures , so that it is not always easy to infer its value in C.G.S. units , but I think the following examples will an idea of the order of magnitude involved .
The quantity of electricity in the form of ions of each kind under normal conditions at the surface of the earth is in electromagnetic measure .
To obtain the conductivity the figures must be multiplied by the ionic velocity per unit fall of potential , the sum of the velocities of both kinds of ions being .
The conductivity of air at the surface of the earth is therefore under normal conditions GERDILN , in one of his balloon ascents , determined the conductivity at a height of 6000 metres and found it to be , which , as far as it goes , confirms the conclusion that the conductivity is inversely proportional to the density .
At a height such that the pressure is one dine per square centimetre , and assuming that the recombination of ions is not materially affected by the low temperature , we should thus get a conductivity of , showing that , if the views discussed in this paper are correct , the ionising power at great altitudes must be considerably greater than that which acts on the air near the surface of the earth .
In speaking of the ionising effects of Rontgen rays , Professor J. J. THOMSON states that even when the ionisation is exceptionally large the proportion of the number of free ions to the number of molecules of the gas is less than 1 to From this I calculate the conductivity to be about at atmospheric pressure .
Some by RUTHERFORD fix the conductivity of air , subject to the action of radium having an activity 1000 times less than pure radium , to be under normal conditions .
These figures would give to the conductivity of air at a pressure equal to that of a millionth atmosphere a magnitude comparable with that required .
We know of much more powerful ionisers than the Rontgen tube or even radium .
An electric discharge itself is sufficient to ionise a gas , as I proved as far back as 'Rapports du Congres International de Physique , ' vol. III .
, p. 438 ( 1900 ) .
'Conductivity of Electricity through Gases , ' p. 256 .
DIURNAL VARIATION OF TERRESTRIAL MAGNETISM .
1887 .
Data supplied by H. A. WILSON* show that in the positive column of a vacuum tube the conductivity reaches the value and the kathode glow is even more highly conducting .
The same author has expel.imented with air ionised in contact with hot platinum , and the data supplied by his diagrams allow us to fix the conductivity of such air as about at a temperature of When the air was charged with the spray from a 1 per cent. solution of a potassium salt , the conductivity rose to , the temperature being The conductivity of a Bunsen burner has been measured by GOLD and found to be .
In view of these figures , which all apply of course to atmospheric pressure , we ought not , I think , to reject the value of as an impossible one for the conductivity of air at high altitudes , but it is necessary to inquire into causes which produce so strong an ionisation .
The increased intensity of the magnetic variation during the summer months suggests directly that we are dealing with a solar action .
This action may be simply an eflect of radiation or it may be due to an injection of ions into the atmosphere .
The former hypothesis is the one which presents itself as the most natural one , though the coronal streamers lend some countenance to the second view , which has often been put forward and sometimes even pressed in support of wildly speculative theories .
Ultra-violet radiation is known to ionise air in contact with metallic surfaces , but the evidence is somewhat conflicting as to the effect of adiation on the air itself Unless the air is absolutely free of dust , the observed action may be due to the illumination of the dust and not of the air .
Dust-free air is so transparent to luminous radiation that it would not be surprising if the ionising effect would disappear , as some experimenters believe it to do , when proper precautions are taken .
On the other hand , Dr. V. SCHUMAN .
has shown that air has a very strong bing power for wave-lengths which are sufficiently short .
Such short wave-lengths are supplied by several metallic sparks , and are freely transmitted through hydrogen .
Nevertheless it seems difficult to believe that , even if emitted by the hottest portion of the sun 's envelope , they are not absorbed again by the surrounding cooler layers .
We are not , therefore , at present in a position to assert that ently short wavelengths can enter the atmosphere and be absorbed in the outer layer , thereby causing iomsation , but we know so little about the conditions of the uppermost layers that we may reasonably retain the view that the powerful ionisation of the air , which we must consider to be an established fact , is a direct effect of solar radiation .
If we turn to the possibility of a direct injection of ions by the sun into our atmosphere , we have to deal with the alternatives of supposing that ions of both kinds are introduced or only those of one kind .
The second alternative must be rejected at once , because a simple calculation shows that the outward force due to the volume 'Phil .
Mag 1900 , p. 512 .
'Phil , Trans vol. 202 , p. 243 ( 1904 ) .
electrification of air necessary to account for the required conductivity would be more than sufficient to overconle gravitation and to drive out the conducting portions at an enormous rate .
The injection , on the other hand , of a sufficient number of ions of both kinds also presents difficulties on account of the large quantity of new matter which would have to accumulate in our air , especially if it is considered that recombination at a rapid rate would take place both in the journey from the sun to the earth and in the passage through the different layers of the atmosphere .
The only alternative to ultra-violet radiation seems , therefore , to lie in the injection of ions travelling with sufficient rapidity to generate other ions by impact .
The air itself , according to this view , would supply the raw material for the ionisation , the injected corpuscles only acting as fertilisers .
There are , of course , other possibilities , such as introduction of radioactive matter , or a spontaneous ionisation which may , if the rate of recombination is slow , be very effective at a great height ; but that the sun undoubtedly plays an important part in the process is shown not only by the summer effect , but also by the periodic changes of the magnetic variation , which corresponds with the sunspot cycle .
I have held for many years and frequently expressed the opinion that the relationship can only be explained satisfactorily on the supposition that the electric conductivity at times of many sunspots is increased .
Whether this is a direct influence of the sun , or only an indication that an ionisin influence is brought into the solar system from outside at times of many sunspots , is a question which everyone is likely to answer according to his individual views of the cause of sunspot variability .
That the increase in the number of sunspots coincides with an increased conductivity of the upper layers of the atmosphere is also indicated by the eleven years ' period of the aurora borealis .
The distinguishing feature of the relationship seems to be this , that auroral displays extend further into moderate latitudes when the solar activity is great .
An increase of conductivity is the simplest and most natural way of accounting for the effect .
The primary cause of the electric discharges which manifest themselves in the aurora is still unknown .
We may Iook for it , perhaps , in electrostatic forces which are always present , but causing a visible discharge only when their intensity rises abnormally , the course and intensity of discharge being much cted by inequalities of conducting power .
On the other hand , there are other electromotive forces of induction not discussed in the paper , such as those accompanying a general drift of the atmosphere from west to east , which may well have something to do with the cause of auroral displays .
Or again , if interplanetary space contains sufficient matter to be conducting , as I believe it must , there will be strong electromotive forces acting in the earth 's magnetic field between the conducting powers rotating with the earth and those of interplanetary space .
Outbreaks of magnetic distul.bances , affecting sometimes the whole of the earth simultaneously , may be explained by sudden local changes of conductivity which may extend through restricted or extensive portions of the atmosphere .
I have shown in DIURNAL VARIATION OF TERRESTRIAL MAGNETISM .
another place that the energy involved in a great magnetic storm is so considerable that we can only think of the earth 's rotational nergy as the source from which it ultimately is drawn .
The earth can only act through its magnetisation in combination with the circulation of the atmosphel.e , so that magnetic storms may be considered to be only highly nified and sudden changes in the intensity of electric currents circulating under the action of electric forces which are always present .
Those ents only are discussed in this communication which produce periodic variations in the magnetic elements , but there are also electromotive forces giving rise to current functions which are expressed by zonal harmomcs and cannot under ordinary circumstances be observed , though any variation of conductivity between summer and winter would produce an annual period .
One further consequence of the theory deserves to be noticed .
The electric currents indicated by our theory are sufficiently large to produce a sensible heating effect in the low-pressure regions through which they circulate .
They will protect , therefore , the outer sheets of the atmosphere from falling to the extremely temperatures which sometimes have been assumed to exist there , and they may help to form the isothermal layer which balloon observations have proved to exist at a height of about 50,000 feet .
Enough has been said to show the importance of the questions on which further investigation of the diurnal variation must give valuable information .
If the fundamental ideas underlying the present enquiry stand the test of further research , we are in possession of a powerful method which will enable us to trace the cosmical causes which affect the ionisation of the upper regions of the atmosphere and which act apparently in sympathy with periodic effects showing themselves on and near the surface of the sun .
It should be our endeavour to put the theory itself to a more accurate test than can at present be done .
The most promising line of attack seems to me to be the investigation of the diurnal variation near the equator , where , as explained in S 6 , it should not only vary with local time , but possess a term depending on the time of the meridian which passes through the magnetic axis .
An exact determination of lunar effects would also , as has already been pointed out , serve as a valuable test of the theory .
PART II .
The problem to be solved may be stated thus : a spherical shell of fluid is animated by a quasi-tidal motion and is under the influence of magnetic forces of which only the vertical components are considered .
It is required to find the magnetic effect of the induced currents if the motion is subject to a velocity potential , where is a surface harmonic , the longitude measured from some standard meridian towards the east , and is the local time of that meridian .
The conductivity of the fluid is not necessarily uniform , but we take it to be expressible in the form , VOL. CCVIII . .
2 here is the colatitude , and measures the difference in longitude between the sun and the place at which is required .
The question is solved if we can determine the current function of the electric currents which are generated by the fluid moving through the magnetic field .
problem for constant conductivity has been treated in the first part of this communication and the interest of a non-uniform conducting power is confined to the case that the variability depends on the angular distance between the sun and the point considered .
If be this angle , the effect of the sun 's radiation will be proportional to in the hemisphere subject to the adiation , .
for values of smaller than .
If the induction is due to the ionising power of the sun 's rays , the rate of recombination of ions has to be considered , but unless this rate is of a different order of magnitude from that observed near the surface of the earth , the conductivity may be considered to be proportional everywhere to the illuminating power .
For values of intermediate between and we must , then , give value to the conductivity .
By means of Fourier series we may now express the conductivity in a series ( 17 ) , which satisfies the condition for , and for Confining ourselves to the first two terms and substituting the value ofcos from ( 13 ) in terms of the hour-angle of the sun and its declination , we obtain @ sin The conductivity has therefore the assumed form if we put ; @ ; S. Were we to adopt the simpler form and put the conductivity proportional to so that it reaches zero value only at midnight , we should have to put @ ; @ , and every case can be expressed in terms of a series such as ( 17 ) , our investigation by proper adjustment of the constants taking account of the first tWo terms .
The term in might be taken into consideration without much difficulty should that become necessary .
The value of can provisionally be put equal to unity and re-introduced at a later stage .
Writing and we may therefore put . . . . .
( 18 ) .
DIURNAL yARIATION OF TERRFSTRIAL In order to avoid frequent interruption , I prove in the first instance a few formulae of transformation which I have found of utility in these investigations .
I start from the following equations denoted in my previous communications by Ronlan letters , which it is convenient to retain : . . . . .
( A ) , sin , . . . . . . . . .
( B ) , , . . . . .
( D ) , ( E ) , .
denotes the tesseral function derived from the zonal harmonic by the relation , where Multiplying ( D ) by and by , and adding , we find , with the help of ( A ) , If in the formula ( A ) we substitute for , it becomes From the last two equations we obtain , by subtraction and substitution of , Now multiplying ( B ) by , and subtracting , we ] obtain .
If on the right-hand side of we substitute the values of and from and , we obtain a corresponding equation ( K ) .
A further useful ansformation is derived from the equations If we subtract and add these equations , they reduce to , and .
We shall require to find the effect of the operation ( 19 ) .
We omit , for the sake of shortness , temporarily the constant , and divide the operation into two parts , the first being From the fundamental equation relating to tesseral harmonics this is equal to - . . .
( 20 ) .
The remaining part of the operation is ( 21 ) .
If ( 20 ) and ( 21 ) are now added , and and are applied , we find the result of the operation to be , restorin . . .
( 22 ) .
We may note that each of the equations used , and therefore the final results , remains true for , if we define .
DIURNAL VARIATION OF TERRESTRIAL MAGNETISM .
This is in agreement with RODRIGUEZ'S theorem , if the definition of depending on the operation is extended to negative values of , for in that case ( 23 ) .
It follows that the operation ( 17 ) , in the case where replaces , reduces to This result may also easily be obtained independently , but in view of the ultimate application of ( 22 ) it is important to include the special case in the general one .
It will be appropriate here to obtain another formula which will be used subsequently .
Let it be required to find ( 24 ) .
From the fundamental equation we find this to be equal to sin and as and ( 24 ) becomes equal to ( 25 ) .
We are now in a position to attack our main problem .
The equations to be solved are ( 14 ) , where X and the values given in ( 1 ) , and for we must substitute the expression ( 18 ) , remembering , however , that we must ultimately restore the factor We may also temporarily omit the factor in the expression for the electric forces .
To find the current function we must eliminate , but owing to the fact that contains and , this does not seem to be possible directly .
The difficulty must be turned by eliminating in the first place , and if is then found , may be determined from the first of the above equations , It may occur to the reader that be more directly obtained if the resistivity were introduced instead of the conductivity .
This is true , but the results are less valuable , as may be seen from the fact that , as suggested above , the Fourier expansion may have to be applied to the conducting power in so far as it depends on the position of the sun .
If the resistivity were introduced as the variable , the high and possibly infinite values which the factor would take when the conductivity sinks low or vanishes would present difficulties much greater than those met with by keeping the conductivity as the variable quantity .
The elimination of requires in the first instance the reduction of i.e. of Introducing the values of X and , the operation reduces to .
The part independent of and is .
The part depending on is We are left with the part dependent on , the factor of which is DIURNAL VARIATION OF TERRESTRIAL MAGNETIShl .
Of this we take separately 2 .
2 .
The remaining terlns depending on are or , making use of and , The terms containing , leaving out the longitude factors , are therefore Collecting our results together , we find as the effect of the operation , eliminating on the left-hand side of equations ( 14 ) , ( 26 ) .
The expression reduces to about half its terms when , and for the two special cases which form the main subject of the present inquiry we have then Case I. .
Case II . .
Our next step must be to find the expression resulting from the terms containing in the elimination of R. We shall begin by assuming to be a spherical harmonic of the form , and shall again take the parts depending on ' separately .
Independently of both these quantities , we have ( 27 ) .
As factor of we have The value of this has been obtained under ( 25 ) .
Finally , as factor of have This is identical with the expression ( 19 ) , the result of the operation being given by ( 22 ) .
If we collect our results , by adding the right-hand side of ( 27 ) to ( 25 ) and ( 22 ) after applying the appropriate factor , we shall have obtained an expression for ( 28 ) .
It will appear that can be expressed in the form of a series ( 29 ) , where is determined by the phase of the velocity potential , which rules the flow of matter in which the electric currents are induced .
We shall avoid the labour involved in the consideration of special cases if we write ( 29 ) in the form ( 30 ) .
DIURNAL VARIATION OF TERRESTRIAL MAGNETISM .
Adopting the ition ( for , where is negative , we may return to the original form by replacing the coefficients with the help of ( 31 ) .
If each term of ( SO ) be subiected separately to the operation ( 28 ) and the results collected , so that all terms depending on any one value are brought together , we may express the result of the operation by a series of the form ( 32 ) , which must be equal to ( 26 ) .
If we put ( 33 ) , we find . . . . .
( 34 ) , ( 35 ) , ( 36 ) .
The values of which may now be found by equating the factors of in ( 33 ) to the corresponding factor of in ( 26 ) , remembering , however , that in the latter equation and have the definite value belonging to the assumed velocity potential .
If , for the sake of clearness , the type and degree of this velocity potential be now denoted by and , we find ' 2 ( 37 ) .
If in any of these values of the index is less than the suffix , that value may be put equal to zero , as the tesseral function to which that coefficient applies is zero .
All other values of are zero .
If we take the case , as an example , we obtain from ( 33 ) , ( 34 ) , ( 35 ) , and ( 36 ) a number of equations in which all values of may be put equal to zero excepting ; the values of these are VOL. CCVIII.\mdash ; A. 2 obtained from ( 37 ) and we thus find equation ( 33 ) for these special values of and to become These and all similar equations in which the right-hand side is equated to zero are cient to determine the coefficients , each in terms of a series proceeding by powers of and .
I proceed to show how the successive approximations may be obtained .
If and ' are both zero , the first of the aboye equations leads to and , as the equation must ] for all values of , this gives us that portion of which is independent of and .
The remaining equations tell us that there can be no other factor which has a term not containing and .
The last three equations contain and they are the only equations out of the complete series which contain this particular factor .
Substituting its value as far as it has been found and neglecting in the brackets all factors except , because they must all depend on or , and therefore introduce quantities of the second order , we obtain a set of three equations which determines those coefficients which involve the first powers of these quantities .
We are thus led to ; ; No other coefficients can contain terms depending on first powers .
If we now write down all equations in which occur , we may determine the terms involving ' and .
Thus if the equation involving is reconsidered , we find that in view of our knowledge just acquired must contain terms in and ' satisfying equation The equations for , give The factors such as , which can only contain powers of and ' higher than the first may be left out of account in solving these equations , and we thus find all terms which contain ' or .
We may proceed in this manner , gradually working by successive approximations from lower to higher powers .
The following two tables contain the results , including all powers as far as the third , for the two typical atmospheric motions represented by the current functions and .
For convenience of use the coefficients now replace the coefficients with negative indices DIURNAL VARIATION OF TERRESTRIAL MAGNETISIsI .
VALUES of .
( Velocity Potential ) VALLTES of .
( Velocity Potential ) VALLTES of .
( Velocity Potential ) VALLTES of .
( Velocity Potential ) VALUES of .
( Velocity Potential ) VALUES of .
( Velocity Potential ) The determination of is only of interest as a stepping stone to the evaluation of R. We must therefore return to the first of equations ( 14 ) , and by its means determine as a series of harmonics in the normal.form .
If we write the velocity potential , we find by means of the formulae of transformation previously introduced ( 37 bis ) .
As , omitting constant factors , is equal to , the tesseral function of type and degree , we may now write and tabulate those values of which are not equal to zero .
The following table is constructed in this way .
DIURNAL VARIATION OF TERRESTRIAL MAGNETISM .
VALUES of The transformation of presents no difficulties .
If be expressed in its series according to ( 30 ) and the terms re-arranged , the result of the operation is of the form , where is up of three parts .
Independently of and we have . . . . .
( 38 ) .
We have further , multiplied by .
( 39 ) , and finally , multiplied by ( 40 ) .
By means of the values of already tabulated , the factors may be calculated .
is expressible in the form or , admitting negative values of , more conveniently by .
In the ultimate result we return to the coefficients through the relation Equating the factors of in ( 14 ) , we now find The calculation of ?
in its present form involves the summation of the expressions ( 3s ) , ( 39 ) , and ( 40 ) , the factors being substituted out of the tables previously given .
The somewhat troublesome labour involved in this process may almost entirely be avoided by a transformation of expression ( 38 ) .
Substituting from ( 34 ) into ( 33 ) , we find and by means of this equation , when and ] are respectively substituted for , we obtain , where If the right-hand side of ( 41 ) replaces ( 38 ) , and is added to ( 39 ) and ( 40 ) , the whole expression reduces to We have , therefore , the following very convenient expression which allows us to calculate the coefficients from the previously established values of ( 42 ) .
The first term on the right-hand side is zero , except for the cases where the type DIURNAL yARIATION OF TERRESTRIAL 1 .
99 does not differ by more than 1 and the degree by not more than 2 the type and degree of the original velocity potential .
For the calculation of these special cases the previous investigation furnishes the necessary formulae .
Confining ourselves to the velocity potentials and , we find the requisite numbers collected in the following table:\mdash ; VALUES of Equation ( 42 ) holds also for negative values of , but when is smaller than it is more convenient to calculate the coefficients from the values of already given .
We find for this case which , as may be expected , is identical with the equation connecting and For , ( 42 ) gives and hence .
( 4.3 ) .
Our equations are not valid for the case , because they depend on a division by .
The first of equations ( 14 ) from which we started containing only in the form is obviously unsuitable to determine those parts of which are independent of ; we must , therefore , have recourse to the second equation Such products as and may give rise to terms independent of only through products of the form , or , sin .
We need therefore only consider that part of which invol es and terms of the first type in and S. Selecting the two terms in the expression for and rejecting all terms not containing , we are left with Similarly By means of the formulae of transformation previously given we find In the special case worked out in the previous pages in which The velocity potential contributes , as far as the term is concerned , nothing to the zonal harmonics in R. We have therefore in the second case , and whenever the velocity potential does not contain a term of the first type , If , as far as its zonal harmonics are concerned , is written in the form we find , as . . . . .
( 44 ) .
We should have obtained the same value if we had applied the general equation to this case .
When , we have to consider the terms depending on , and must therefore write For all other values of equation ( 44 ) applies .
DIURNAL VARIATION OF TERRESTRIAL MAGNETISM .
The tables which follow give the calculated values of and , and therefore solve the problem as far as terms of the fourth order in and .
It will be noticed that for unity , which is the highest admissible value of and , all the factors involving higher powers than the first are so small that their eflect falls much below anything that observation is capable of .
Hence the approximate calculation given in Part I. is sufficient for all practical purposes .
Restoring constant factors , we may summarise the result of the previous investigation as follows:\mdash ; 1 .
Notation .
denotes the tesseral function , where is the zonal harmonic of degree , and the colatitude .
measured upwards denotes the vertical magnetic force of the earth 's permanent field at the geographical pole .
is the thickness of the conducting atmospheric shell .
is the conductivity , which is supposed to be variable and depending on and the local time , according to the relation sin where and are constant .
is the velocity potential of the flow of air .
2 .
Conclusion .
The current function of electric flow induced under the action of the vertical force in the oscillating shell of air is then expressed as a sum In order to obtain the netic potential of the variation caused by the flow of air , a factor has to be applied .
The factors and are giyen in the tables ( including terms of the fourth order of and ) for the velocity potentials and VOL. CCVI[I.\mdash ; A. 2 I. VELOCITY Potential : DIURNAL VARIATION OF TERRESTRIAL MAGNETISM .
II .
VELOCITY Potential : II .
VELOCITY Potential : ( continued ) .

