MARCH, 1916. MONTHLY WEATHER REVIEW. 115
I- I I
p 9. " (, /I/ e:,
'j -
.I / SECTION IL--GENERAL METEOROLOGY.
SOME PBOBLEMS OF ATMOSPHEBIC ELECTBICITY.
Ry GEORQE C. SIYPSON, D. Sc., F. R. S., vice president.
(Read before the Physics Section, Indian Sc~ence Congress, Lucknow, January, 1910.)
[Dated: Meteorological Department, Simla, Indla, Feb. I, 1916; received, Mar. 31,1816.1
The problems of atmospheric electricity with which we are about to deal must be solved by nieans of the laws which govern the behavior of electricity in y e s , and I therefore propose, before discussing the actuit conditions, to sketch in a few words the main rinciples involved.
atmospheric electricity may be considered to be a per- fect conductor. Its dimensions are so large, and there- fore its electrical capacit,y so peat, t,hnb no aiiiounts of electricity which we can extract or add nt any place can alter its electrical potential. As the pntent,inl of the earth does not change, it is convenient, to take its poten- tial as our standiwd and refer d l other potentials to it. We therefore describe the earth as being ab zero poten- tial and the potentials of all other bodies as plus or minus the difference between the potential of thc body and the earth. Let us suppose thst t,he earth has no atniosphere, and
we give to i t a chtrge of electricity which, in order to fis our attention, we will assume to be negative, then by the laws of electrostatics the charge woulcl distribute itself all over the surface. If the surface were a perfect sphere, the surface distribution would be everywhere the same; but as there are irregularities, every hill woiilcl have an escess of electricity and every alley a defect. The value of the surface density a t any point, could easily be measured. An ideal, although not a pract,icd9 method to do this would be to take a unit chnrgo of
positive electricity and hold it a t a c.ertn.in distance froni the earth. The negative charge on the eart.li would then attract our positive charge, niid if wo movcd thc lntter away from the earth's surface we ~hould havc? to do
work. If we moved the charge from the ground t.o :I
meter above the ground. we should do against the elec- trical forces an aiiiount of work equivalent to the electri- cal potentid energy of the charge a t 1 meter above the ground. 'More work would have to be done to convey the charge to 2 meters, and still more to 3 meters. At
everv position, then, above the surface our unit charge would have a different potential. E'roni this esperiment we could calculate the change in potentmid of our unit charge a t every position above the surface,, and it would be a simple matter, by the laws of electrostatic.s, to cal- culate from this change of potential, or, as it is called, the potential gradient, the amount of charge on the earth's surface. It is important to realize that! when wc measure the potential gradient, immediately above t,he earth's surface we are only finding the force due to the electricity spread over the ground. Now, let us imagine that an atmosphere is given to the earth, and see what consequences we might espect. For long it was supposed that, air was a perfect noncon- duc.tor of electricity, and we .wiU for a short time retain this false impression. A nonconducting atmosphere a t rest would not affect the electrical state of the surface
The earth is a solid body whici P for all purposes of
and we should still be able to measure the amount of charge by measuring the potentid gradient. If, how- ever, the air moves and we get winds, the electrical state of the surface ma be greatly affected. I have explained how the whole c 5 arge on the earth is situated on the upper surface of the ground; hence ever particle of
char e of electricity upon its surface. If, therefore, the
tricity produced by the process) or w rls away the leaves of a tree, a certain amount of electricity is sepa- rated from the earth and raised into the air. But so long as the air is nonconducting the charge remains fixed onto the dust and onto the leaves, so that when the wind stops and they fall to the round again the charge is
returned to the surfme of t a e earth. With a noncon- ducting atmosphere the charge on the earth can not be periiianently separated, and after any len th of time we should still find the same quantity of e 7 ectricity from our measurements of the potential gradient. But recent research has shown that air is not a erfect nonconrluct,or; it conducts electricity slowly, Rut as
%
surely 21s r o ,per conducts it. Electricdy R conductin atmosphere Ibecomes a part of a coiiduc.ting eart,h, an 'ust as before there was an atmosphere electrostatic iorces drove all the electricity up to its surface, the same
forces will drive i t through the air unbil it reaches the surface of the conducting sphere which is now the con- fines of the conducting atmosphere. All our electricity then will ult,irnat,eJy spread itself in a uniform la er over the outside of the nt'mos here. Now, it is we1 f known that there is no electrica P force within a conductor no matt,er how much electricity there may be on the sur- face. Therefore no experiment's that' we can make at tlie surface of the eart,h would reveal the charge spread over tlie upper atmos here. I t is important to realize t.lint t,liere niny be a cfarge of untold amount in the up- per i~bmos here of which we are entirely ignorant .be-
no elect,rical force within itself. I t therefore appears that it would be impossible for a charge to remain on the surface of the earth while the air is ever present to conduct it away into the upper at-
nios here. Yet we find that the whole surface of the
ishino unnt,ity of electricity. This is a parados about whici ? shdl have a great deal to say later, but before discussing it, it will be necessary to go more fully into the causes of t,he conductivity of the air. The coiduc.tivity of the air is a very variable quantity, its average value is near to 3 x io-' electrostatic units, which means that it,s resistivity is 3 x 10 ' ohnis or twenty thousand million (3 x 10 lo) times that of cop er. The meclinnimi of the conductivity in copper an8 air are,
however, very different. The air conducts only when its molecules are ionized: that is, split up into positive and negative ions which move under the influence of the electrical field. Thus, when a current of electricity passes through conducting air there is an actual transferrence of matter, while in cop er the elec.tricity moves independ- ently of the mass of t P ie copper.
dust, every blade of grass, and every leaf o 9 a tree has a
win B raises the dust (we are to neglect an frictional elec- li
cause of t P lis property of a conclucting body to eshibit
e& r 1 is charged and reiiiains charged with an undimin-
116 MONTHLY .WEATHER REVIEW. MARCH, 1916
................................................ ................................................ ................................................
a
i3
- ,
1. RADIOACTIVE SUBSTANCES AND ATMOSPHERIC
IONIZATION.
Observations in a.ll parts of the world have shown that the atmosphere everywhere is ionized and it is enerally
are the cause of this ionization. The first question which I wish to discuss is whet,her this radioactive t,lieory of the ionization of the air is alone capable of explaining the facts. To do this we will first consider the ionization ac.tuaUy found by experiment over the land and over the sea, then the number of ions formed by t,he liliown radio- active matter over the land and over the sea, and lastlv whether these values agree. "he ionization, over the land particularlv, is greatly affected by the meteorological condibions, the chief fac- tor being the temperature. We s h d therefore in the following only consider the ionization with tenipera.t,ures between 10' and 20°C. Very many measurements of the ionization have been made and these show t.hnt within the limits of tmnpernture we are considerin-, t,here are on the average over the land 1,000 ions of ea&
si n in each cubic centimeter of air. %he measurements over the sea n.re much less numer- ous, but when on a voyage froni Enoland to New Zedand Wright and I 1 made a number of %eterniinntions of the ions present over the ocean far from land, and the mean vnliie of sist,een obsemat.ions wit,li t.he tmiperature between loo and SOo C. was SO0 ions per cubic centiinet.er. We therefore have over land 1,000 and over the ocean far from land SO0 ions per cubic centimeter. We will now turn our attention to the radioactive mat- ter which is supposed to be capable of producing this ionization. It is now definitely known that the rocks of the earth's crust contain appreciable amounts of radium and thorium, and numerous measurenien ts of the radio- active contents of the rocks have shown that this radio- active matker is present, in uenrlg all kinds of rocks and fairly uniformly distributed throughout all soils. The
radioactive mat,ter in the ground produces a, 8, and 7
rays, but the two forn1.w do not escape into the atnios- phere, exce t in a negligibly snisll proportion from the actual s d c e , and therefore cannot effect ionization there. On the other hand, owing to their greater pene- ower, n certain roportion of the y rays do pass
are absorbed in the ctir and therefore the ionization due to them decreases as we ascend but iviiig our attention
calculation3 shows that the aver e amount of radium in the soil ves out sufficient 7 ra iation to produce 0.S
thoriuni is not known with such accuracy, but it is gen- erally su posed that the ionizing ower of the thorium
the thorium and ra ium in the soil combined produoe by means of their y radiation something like 1.6 ions per cubic centimeter per second in the air just above the ground. The radium and thorium in t,he soil are constantly giving off their respective emanations, These emana-
tions fill the interstices of the soil, from which they escape into the atmos here by ordinary diffusion and in conse-
held that the radioactive substances in the eart % and air
tratil? out o t R e ground into t t: e air and ionize it. These rnys
to the air near the surface, say, wit f iin 6 feet,, a simple
ion per cu f- ic centimeter per second. The amount of
in the so1 is about e ual to that o P the radiuln. He1ic.e !i (g.
quence of any B all of bakometric pressure. Thus the air
lonslcc., scc. lonslcc., see. Iona/cc., ace.
1.00 a63
0.06
1.63
0.035 0.035
0.035 0.025 0.06
1 Simpson & Wright in Roc. Roy. soc., London, 1911, MA: 175.
2 Eve In PhIl. mag., London, 1911,21:26.
...................................... Total 1 1.70
I
~
over the land receives emanations from the ground which behaves very much like one of the ordinary gases of the atmosphere esc.ept that it is constantly decaying owing to radioactive change. This enianation or radioactive gas is constantly emit- ting ?, 8: and radiations and in consequence ionizing the air. In this rase all three kinds of radiation are in a position to ionize the air and we are able to forni some
idea of the magnitude of the effect. The calculations of Eve show that the actual radiation from each of the emanations in tho air produces the ions given in Table 1.
TAULE l.-Inns per cubic! cetitimatn pw sccond produced in the atmosphere by the diffcrent cniunutioxa in fhe air.
Rays. I Radium. 1 Thorium. 1 Total.
1.05 a. 75
Sources of rays.
Ions/ce., aec. lons/cc., see.lIm/ce.. 8m.
Air .................................... ......... [ 1.701 1.051 2 7 5
............................................ 0.80 0. BD 1. w) Emth
Told.. .................................... I 2.50 I 1.85 1 4.35
I
Thus t,lie best estimites show that in each cubic ceuti- meter of air over the laiid there are 4.35 ions of each sign geiierated every seconcl.
Siniiliw ronsiderrttioiis can be applied to find the num- ber of ions generat,ed b,y radioactive iiintter over the ocean. We have tho radioactive matter in the sea and the ern%natio!i in the air. The radiosctive matter in the sea itself' is so snisll t4hat it cxii not produce by means of
its y rays more tlim 0.01 ion per cubic ceiitinieter per
soc.oiit1 in die air over the sea. Also the eninnation given out from the sea is so minute that it can be neglected, therefore if there is aay eniaiiatioii in the air over the sea
i t niust heve been brought from tho 1md by the winds. This niitkes it inipossihle for tliere to be any appreciable t,horium enianntioii .over the oceans, for thorium eman?- tioii reduces to half value in lcss than a minute and It would therefore have en tirely clisappearetl before the wind could have carried it far from the land. We are, therefore, left with radium emanntioii OilIy in the air over the ,son. Naturally the amount of rncllum
emanation over the sea has not been determined to any- thing like the same extent PS it has over the land. But we know that it is very much less. At Hammerfest, in
Norway, I found a that whsii the wind blew from the west, i. e., from the' Atlantic Oceau, the radium it contained was only a little more than a twentieth art of the amount brought from inland by southerly Win f 3.
3 Simpson in Phil. trans., Royal m., 1905,205A: 61.
MARCH, 1916. MONTHLY WEATHER REVIEW. 117
\ row!c. c.. arc.
sea. ........................................ I 0.01
Tot d... .............................. I 0.1s
Air... ...................................... 0.17
During the voyage already referred to Wright and I measured tbe radioactivity of the air over the Atlantic and southern Indian Oceans far from h i d . Wlitxi WG
arrived at South Africa the ap aratus we had used 011
the r J ium collected inland was twenty tinies that col-
lected over the ocean. These ex eriments indicate, then, that over the or em^
over the land, but as other obsorvers have found slightly higher values we will assiinie that there is 10 per cent. Thus, the radium in the air over the ocean can produce at the most only a tenth of the ions produced by the radium over the Isnd. We therefore are justified in ronstructiop for tslie ocean the following table to compare with Table 2 for the land.
TABLE 3.-Ions per mbic centimeter prr serond prodwcecl i n the air ncnr thc szirfnrc of the scu by the radium and tliol.izcin in both air mid sea.
the shi WAS taken 200 miles in P and. I t wis found that
there is 011 P y 5 per cent of the radium emanation found
--
Source 01 rays. 1 Radium. I Thorium. I Total.
ronsic. e., scc. 1 rmwjr. e.. wr. 0 0.17 0 1 0.01
0 I 0.14
(rrerland ............................................ Overocean.. .........................................
Localltg.
per eubiccenti- cuhir centi- rnet.er per meter. I second by dioactiw
i matter.
(?ti (B)
I-. ......
1,m
~
4.35
so0
~
0.1s
We must now esamine whether the nunibcr of ions
generated in each case is capable of niaintaining tho number of ions ac,tually found. If every ion formed in the atmosphere remained an ion we should have an ever increasing-number; but w e know that when a positfive ancl a negative ion meet they rcrcombine to form a neutral molecule. It is obvious that the more ions there m in a given volume tlie more rapidly they will join together, and i t is easily shown that for the steady state we have the folowing relation- ship :
p-mz = 0
in which q is the numbor of ions of each sign formed each second, n. t,he number of ions of each si n present and
a the coefficient of recombination. %lis equation simply means that for the stead state the number of ions formed in a second, q, is equay to the number of ions which recombine in R second. If we can determine a for the land and the sea we
shall be able to decide ~het~lier. the values of n and p
given in Table 4 are consistent. The rate of recombina-
tion of ions has been determined by several observers. The method used has been to ionise air by some outside ioniser, X rays or Becquerel rays, then to remove the ioiiiser and determine the decrease in ions due to recom- bination after definite intervals of time. Before discussing the results of these measurements, we must say something about the factors on which the rate of recombination de ends. If an ion comes ne.= to matter, say the wall o the vessel containin the am, it induces a charge on the wall and the electrica force so
produced causes the ion to attach itself to the wall and it
is lost. In the same way ions attach themselves to any dust floatsing in the air and in conse uence lose their property of moving freely in an electric %l eld; agam these are lost, to measurement. It is found that water vapor tends to attach itself to ions and in consequence the ions lose their mobility in a damp atmosphere. Thus, dust and damp in t,he air have the apparent effect of increasing the rate of reconibination of ions, hence we should expect the rate of recombination to be least in dry, dust-free air, greater in damp air, and greatest in damp and dusty air. Esperiments hare confirmed this conclusion, and the values of a found hare been- (a) in dust-free dry air, 1.5 x lo-" (Townsend ').
(b ) in the sir in the outskirts of Vienna, 3 x lo4 (Mache & Rimmer').
(c) in the dusty and damp winter air of Manchester,
4 x 10-6 (Schuster O).
In order to nrrivc. at a definite conclusion, we ought t.0 have ninde our deterininations of n., p, and a simul- t,aneously. As, however, this is impossible, we must choose from the above values of a the most probable value for the conditions under discussion. We can say at once that a both over the sea and land must be greater than the value found for dry dust-free air. As the values of n for the land given above were det8ermined in pure country air, we can say equally defi- nitely that a! must be less than the value found during winter in Manchester. Also, it is probable that a is greater over the land than over the sea. It therefore seems reasonable to take a over the land
tis 3 :.: With t,liese values of a and the rates of generation of ions given in Table 4, we find from the equation
T P
ancl a over the sea as 2 x lo4.
.-
that, if the whole ionisation were due to the radioactive matter known to be present, the air over the land should contain 1,300 ions per cubic. centimeter, and the air over the sea only 300 ions per cubic centimeter.
Coni aring these numbers with those given ip the 880
over the land is quite sufficient to aecount for the ionisa- tion found there; but this is f a r from being the case over t,he ocean, where there are nearly three t h e ? as many ions as can be accounted for by the radioactive matter present,. It therefore appears that over the ocean at least the radioactive theor of the ionisation of the air is not satis-
I wish to bring forward for discussion.
ond co P unin of Table 4, we see that the radioactive matter
fackory. This, t t en, is the first unsolved problem that
4 Towmend in Phll trans R 1 Societ 1888 A M : 157.
8 Schuster in Mem.. Mancieater lltt. Ad Ph. nw., 1904.48, ?dam. 12. Marhe br Rlmmer'fn Ph.'aik%cbr lb 7:h17.
118 MONTHLY WEATHER REVIEW. MARCH, 1916
2. EARTH'S PENETRATING RADIATION.
M second problem is generally described as the prob em of the earth's penetratin radiation.' If a closed metal box of suibab e dimensions is cleaned with the greatest care so as to remove a.11 radioactive matter from the inside, and if it is filIed with air from which all radioactive matter has been removed by pmsing it through liquid air, it is found that the air within still has a residual ionization. This in itself would not be very surprising, for it might be due to some small radioactivity of an im urity in the
property of all matter. What is surprising is that even when the box is ke t at a c.onstant temperature and her-
changes both when the box is moved from place to p1a.r.e and also when it is kept in one place withoutr being clis- turbed. It is obvious that such c.lia.nges must he clue t,o a radia- tion entering the box from the outside, and as the walls are always too thick to ttllow eit,her CY or /3 rudintion t.o
enter, the radiation must he of the y type. We know of only two sources of y radiation: The
radioactive matter in the ground and the radioactive emanation in the atmosphere. The latter, howevrr, can be ruled out at once for, as we see from Tablo 1, it can produce at the most only 0.06 ion per cubic centimeter per second, which is a umtity far too small to measure.
the ground are able by their y rays to produce 1.6 ions
er cubic centimeter per second, which aniount may be ely exceeded in places where the ground is unusually
Neglecting, then, the ionization which is characteristic of the instrument and can not vary, we h o w of only one source of radiation which enters the instrument from the outside and can vary from place to place and time to time. This source is the radionctive ma.tter in tlie rocks and 'soil. The 7 radiation from the rocks and soil although relatively very penetrating has an ap Ireciable absorption
water and is rapidly absorbed by the air. Eve * has calcu- lated that at ltn elevation of 100 meters the radiatioii from the ground shoulcl be diminished by 36 per cent and it should have disappea.red entirely tit an elevnt.ioii of 1,000 meters. One would conclude from this, therefore, t h t i f we
measure the ionization within our vessel over the land, then remove it to a place far removed from rocks and soil, as, for example, to a place over deep water or to a lace 1,000 meters in the air, the ionization within the !ox would decrease by the amount due to the radiation from the earth. Also as the remaining ionization would then bo due to thekstrument itself, one would c,onclutle that it could not be decreased further. Such experiments have been made by many observers, but not with this expected result. When the apparatus hm been removed from the lantl over deep radium-free water a decrease in the ionization has been found, the average decrease bein about 3 ions
ahodd expect. If the apparatus is now sunk into the
m&&?ol. de Frerrca. parls, Octrnov
f s
walls of the box, or even to some genera F radioactive
metically sealed, t R e ionimtion within undergoes large
On the other han 8 tlie known radium a,nd thorium in
in radioactive matter.
coefficient. It IS entirely cut off by I ess than 1 meter of
per cubic centimeter per second which is o 7 the order we
1 A good Mumd of thh subJeet, up to 1912, is givm by Ckuuwu In Annualre, SOC. 1912.
Ew in Phil. msg.. Londtm, 191l~ZI: 28.
water, a further unexpected decrease takes place, the new decrease (2 ions per cubic centimeter per second) being only slightly less than the previous one.# It appears then that, by sinking the apparatus into the water we are cut- ting off another radiation which can only come from above, and is nenrlg as st,rong 118 that due to the radio- active mat.tm in the soil. Similar esperiments h:ive been made on land by build- ing screens of lead about the apparatus, and these have also shown that, some radiation apparently comes from above as well
The results attained by taking the apparatus up into the air are also important. As has been alread pointed out,, if the penet,rating raclintion came only s rom the ground i t would be rapid cut off by the air and at 100
meters it would be reducec i by 36 per cent. Many observations have heen inade on towers and all have shown that the decrease of ionization within the tti>p:artlt.us is much less thrtn the theory re uires. A typical e x a q l e is ti set, of ohservstions made x y Wulf I1
a.tr 300 met,ers on the Eiffel Tower. The ionization was only recluc.er1 by 40 )er cent., wli!le according to Eve's calculat~ions it should h ave been reduced by inore than 90
per cent.
Such obserrstions are, however, not satisfactory, as the towers may possibly have a considerable deposit of radioactive mnt.ter upon thein \.\.hich might be the cause
of some increwe of ionization. 0hserva.tions in free balloons, however, are free from this objection, and quite a number of these have heen
~i i a d e .~~ They d agree in showing a much smaller de- crease in the enetrating radiation with aqcent than would be 'ven radiation from the ground only.
The b 0011 o servations, however, go much further; they show that after about 2,000 meters the decrease
with ascent ceases and the radiation commences-to in- crease. At first the incrense is slow and at 3,000 meters the ionization has returned to tlie value found on the This increii,se wit,h tlie height above sea level
F1 ias also hoen found by Gockcl la to occur when observa- t,ioiis are made on glaciers in the Alps. There can, there-
fore, be little doubt as to the reality of the effect. As one
ttsscends still higher a. strange phenoinenon is observed :
The rate of increase of the radintion goes up by leaps and
bounds. At 9,000 meters there are protluccd each second 9 ions more than on the ground; at 5,000 meters 19, and tit, 6,000 iiieters 30.'' It sppews from aU these observa- tions t.1ici.t in addition to the y radiation from the earth there is anothcr rticlintion conling from the sky. Sinking the apparatus iii water shows that at sea level t.he sky ritdiation can produce 2 ions per cubic centimeter per second and is t.liere€ ore near1 equal to that from the rocks and soil. Further, the g alloon observntioiis show that this radint,ion increases with height. At first, how- ever, as one ascends over tlie land the total ionization within the box decrenses owing t,o tlie cutt,ing off of the radiation from the ground, so that at 2,000 meters the ionization is somewhat less than on tlie ground; a-bove this the radiation incremes rapidly and at 6,000 meters it. is
froni tlie ground.'O
.
9 Goeliel in P
10 Cooke fn Ph?
11 Wulf In Physikal. i!tnchr., 1610, li: Sll. Also sea last pardgraph 01 Eve in Phil.
1% ~B S S in Wlen. Ber., 1912. 121: Wl. Kolh6rater in I'h dbal Ztschr., 1913. IJ: 1066,1153.
11 Gockel in Phydkal. Ztachr., 1915, IC 315.
I( I sea from a reference In a recent paper b Gockel that 80 ions per second have been obwved'at 9.OOO meters: but I have not yefbe8n able to me the original communies- tion.-Q . C. 5. [CI. Abstract of Kolhllrster's paper, YONTELY WEATHER REVIEW, Dec.
sikal. Ztschr., 1915, 16: 34.5.
London 1903 6: 403.
mag. IOC. clt
1915, qj: 596.-ED.l
MARCH, 1916. MONTHLY WEATHER REVIEW. 119
more than ten times as much as that due to the radio- active matter in the ground, and still greater values are found at greater heights. Two sug-
&ions have been rnt-i.de. First, that there is an un- Enown radioactive gas in the air which is niainly confined to the up er atmosphere. Second, that a true pene- trating ra (P iation enters the atmosphere froin cosmical space. It would take me too long to discuss all the pros and cons of these suggestmions. It is sufficient, to say that
both are hi hly improbable on our present knowledge,
Here is a field for research which holds out weat ossibilities; but I am afraid that, i t is one which win not !e investigated further at present. Germany u to tlie present has been practically the only country wkch lins made atmospheric-electricity observations from bnlloons, for in recent years ballooning has been n populttr sport in that c.0untr-y. This sport was no doubt fosbered by its military associations, and i t is very unlikely that, it! will survive the great war. Ballooning is an espensive pas- time and i t 1s unlikely that after the ww my Euro can nation will have money to spare for the purpose. &din is unsuitable for ballooning, so we have practically only America to look to for the investigation of this interesting question. Let us hope that observation will soon he
inade in that country. This, then, is our second problcm: What, is the origin of the penet.rating radiation which protluces 3 ions pi?r
cubic. centimeter per second within a closed box at Sei&
level, and very inany more as cine iisceiiils into the upper atmosphere 1
What can be the source of this radiationP
although per 5 aps not impossible.
3. THE ORIGIN AND MAINTENANCE OF THE EARTH'S
CHARGE.
Wc must now leave the subject of the ionization of the air, its magnitude and cause, which has giren us two important unsolved problems and return to our consid- eration of the charge on the earth which was the starting point of our discussion. The problem that I intend to consider nest is the origin of tlie charge on the grouiid and how it is maintained. It has already been statcd that the charge on the earth is measured by the potential gradient that it produces, and observations of the potential gradient have shown that during fine weather every part of the earth's suifttco? sect, ltmd, plain, and mountain, from north polar regions to south polar regions, has a nemative charge w1iic.h is fairly constant except where the sfinpe of the land causes
local excesses or defects. The potential gradient, has also hcwi measured in t.110 upper atnios here,I5 and it has been found that it de-
only a tenth of what it is near the ground. in the field in n vertical direction can only be due t.0 t e
air cont,aining free posit,ivc elect8ricrity, whic.h connter- balances bhe neg&tive char e on the ground. The potmen-
tial gradient, conbinues to %ecrsase slightly to the highest altitudes reached, and if, as is generally assumed, it, dis- a pears entirely a t great altitudes this can only iiiemi
t%t tlie whole of the ositive charge corresponding to the
The natural inference is that the electrical field in the atmosphere is due to soxe process which has separated
w Linke in Abhsndl. d. mil. Geaellach. d. Wissensch. w Gattinpen, lW, 111, No. 5.
creases rapid P y, so that nt about 3,000 nieteiB the field is
This challi?
negative charge on t. R e surface is contained in the air.
- ____--__ ___.___ -____-
the negative electricity on the surface from the positive in tho air. It is not sufficient for this process to have caused the separation once for all, for on awcounb of the air being a
conduckor a current of t?le,ctricity is sat up betmween the charge on the earth nnd the chrge in the air which tends to ni!utralize thc chnrws and to cause the field due to
them to disappear. TRat this current actually exists c m
be shown by insulating a port.ion of the earth's surface and measuring the actual timount of electricit,y which leaves
it each mc~nd.'~ It is then found that the loss from the sui-fwe is exactly the amount which is calculated for the current caused by the potential grtidient and conductivity
of the air. The loss which thus constctntly tnkev place on account
of the conductivity of the atmosphere must be constantly replaced, and ns long as we hold thst the positive charge in the air itlid the chargo on t h ground are complemen- tary the proc.ess which niin t.ains thc charges and field must, bc? sought, mainly in the lower atmosphere, where the large proportion of the seperated charge exists. Numerous sug estions have been made to esplain the
izing cuiwmt, and they can a1 be divided into two classes: (a) Suggestions which suppose the electricity to be
separated in tho air and tlie ncgative electricity to be carried by some mechruiical means to t,he ground, leaving the positive charge in the air; and
(11) Suggcst,ims which suppose the sepctrat.ion to take pkce nb t,he cwth's surf~ice, which retains the negative diarge while the positive charge is carried to the upper atmosphere in ascending air currents. I think it is worth while to give an esaniple from each of these olasses.
"lie fiist and niost important theory is generally called the Wilson-Gerdien ts11e0ry.17 According to this theory negative ions are nuclei, on which water vapor is readily deposited; hence whcn it! rains the negative ions are car- ried to the ground with the rtiin drops. Thus every rain shower has been lookccl upon 8s carrying negative elec- tricity from the air to the ground, and so maintaining thh; electrical field in t:he atmosphero. This theory received a fatal blow when it was found that rain in d l parts of the world carries down more posi- tive than nogabive olect.ric?it,y. Ebert's t,lieory is n good csaiiiple of the second cl:1~ss,'~ and deserves to be specially iiientioned RS it is still seriously maintained by n large proportion of German physicists. Experiment shows that when ionized air is passed through conducting tubes, the nir under certain conditions emerges with more ositive than negatire ions. Applyin% this to t,he ear& Ebert snys tshatg the int.erstices in t e soil m e all full of radium emanation; hence the air in the soil inustI be highly ionized. When tshe baromoter falls, this hi hl ionized as streama into
the soil, which are equivalent. to the tubes used in the lahoratorv. Hence the air will emerge with a charge of positive electricity which is ra idly disseminated through- out t,he atmosphere by a.scen&ng air mrren ts.
but it does not appear to have convinced Ebert's disciples.
P
maintenance of t i e 7 ficld in s ite of the constant aeutral-
the atmosphere through a1 Hi? t e channes Q and cracks in
The fa.Uac.y of this reasoning has been poiiited
-
18 Wilson, C. T. R.. in Proe., Royal 8oc.. 1908. SOA: 537. Simpson. Geor e C in Phil. mag., London, 1910, 19: 715.
17 Gerdien in F%yiirital. ztschr 1% 6- 617
Sim on in Phil. ma London 19oi) ikbi9.
1s &rt in Phvsikal.%schr.. id. 5: i s Meteorol. Ztschr.. 1 ~. 21: 201.
I@ Simpson in %'h dkal Ztacbr 1904 5' 325. 734.
OW- in Ph h ztwh ih d ~4 . Ebert i n Phy& Ztsehr., ?W, d: &Q; 1905,6 825,828.
&RCH, 1916 120 MONTHLY WEATHER REVIEW.
The interstices in the soil are so ver small that in spite
ions are absorbed b the walls at once. Also the rate of
barometer IS so small that the air emerges with neither positive nor negative ions, for the show that the ionized air must travel through the tubes if the air is not to
charge in the passage. This is not the place to go int.0 details of what has proved to be a long and unsatisfactory controversy. I shall therefore content myself b stating t,hat no process
charge in the air from the charge on the ground. I now propose to show that the charge in the air is not the charge which has been separated froin the electricity on the surface, and therefore it is useless to look for a process which constant1 effects such B separation.
in accordance with the current'which can he cnlculated from a knowledge of tho potential gradient rtncl t.he coii- ductivity of the air just abovr: the surface. Measure- ments have been made of bot-h these f actois b Gerdim ao
that in 31 heights of the atmosphere the potential radient
words the same vertical current is resent throughout the
carrying measuring instruments. Now, as the same vertical current is resent in 6,000
meters a3 on the
with the positive electricity in t i e air. This leads to the important conclusion that t.lie negat.ive chn.rge on the surface and the positive charge in the air are not comple- mentary in the sense that one has been estracted from the other. The relationship between these two charges
of the emanation there can be no hig g ionization, for the
flon through the c K annels and cracks due to ti falling
has been suggested to account 9 or the separation of the
I have already state 2 that the loss from the surface is
to a hei ht of 6,000 feet (1,82S.S nieteisj, an K it is found
multiplied by the conductivity is a consttint. P n other
atmosphere up to the greatest heig \ t reached by balloons
ound, this ineans t l a t none of the negative charge w T ich left the round has combined
'
P
tain time the accumulated charge on the earth has risen until i t has set up, in the atmosphere, at field sufficiently reat tozany away the char e as rapidly as it is received.
earth receives its constant sup ly of elec.tricity. When
us1 quantities of electricity passing upward in ~1 1 parts
2 the atmosphere. If the coaductivity of the air were the same throughout the atmosphere it would need the same force to drive the same current through all layers. But the air does not have the same conductivit through-
!Ius is a final steady state w Fi; ich continues !omlong as the
this state has been reached t g roughout there must be
out, the conductivity increases with height an t f therefore
10 Gerdien in Nachr. d. k. Geaelhh. d. Wis#M. zu Gettlngen, 1805. Heft 5.
the force necessary to drive the current decreases with hei ht; in other words, tho potential grsdient decreases
ductivity so that a constant current can flow, can o 11y be effected by the accumulation of B volume charge of elec- tricity.
The following is a statenlent of the process by which
the steady state is reached. The charge on the earth itself incrertses until it produces such a field in the air
immediately above the ground that the electricity is con- ducted from the surface at the rate a t which it is received. Now, in another layer some distance above the ground the conductivity is greater, hence if the fiald due to the charge on the earth extended so high it. would produce too large a current through the layer; in other words., the lnyer would lose electricity faster than i t received it from below. The consequence would be that iincoinpensated positive electricity would appear in the layer. This would counterxt some of tbe field due to the negative charge on the round and the field i n the layer would decrease.
so induced, woulcl be cut dowx until a stead state was reached in which the current from the grounJwould just be condwted through it and no more. This process would go on throughout the at#mosphere; in every layer sufficient free positive charge would a pear to reduce the
through it.
Thus the two independent varishles are, a, the rate at which t,he earth receives its rharge and, b, the ronduc- tivity of the atmosphere. Given these the potential gradient and the volume charge adjust themselves until the same quantity of electricity passes through every layer of the atmosphere. But these are exactly the conditions which we find in the earth's atmos here; the charge on the earth, the volume charge in t l e air and tha potentia! gradient are so adjusted that the esisting conductivity is 'ust sufficient
the atmosphere. We inust therefore conclude that the
prime cause of the electrical field in the atmosphere is that the earth receives n constant charge of ne ative
wit % height. This adjustment of the force to the con-
Fim f ly the field in the layer owing to the positive charge
field to the value required to coaduct t 1 e constant current
to carry a constznt vertical current throug i all lsyers of
electricity which must be conducted away as rapi 5 ly rn
it is received. We can go further and say that as observations have shown thaz the same currcht passes the 6,000-meters layer as leaves the ground, the supply of electricity to the around can not be obtained from the atmosphere below -
~
this height.
We now see whv all attemnts to solve the problem of the earth's ne at& charge hfconsideria proc&ses which
failure.
To solve our problem then we have to discover some means by which the eart,h may receive a constant supply of negative electricity without the corresponding positive charge being set free within the earth itself or the lower
are mnipletef wit,hin the lower atnlosp fa ere have led to
atmosplpre. This 15 a much more difficult problem and one which, it may be stated a t once, bafAes all known physical pro- cesses. When lon and serious attem ts have been made to
appears to me that it IS justifiable to draw on one's imagi- nation and to atate what kind of rocess robable or not,
would satisfy the conditions. 5 therigre pro ose to
solve a prob K em along recognize cp lines without result, it
describe two proceases which would explain the p i enom-
&RCH, 1916 120 MONTHLY WEATHER REVIEW.
The interstices in the soil are so ver small that in spite
ions are absorbed b the walls at once. Also the rate of
barometer IS so small that the air emerges with neither positive nor negative ions, for the show that the ionized air must travel through the tubes if the air is not to
charge in the passage. This is not the place to go int.0 details of what has proved to be a long and unsatisfactory controversy. I shall therefore content myself b stating t,hat no process
charge in the air from the charge on the ground. I now propose to show that the charge in the air is not the charge which has been separated froin the electricity on the surface, and therefore it is useless to look for a process which constant1 effects such B separation.
in accordance with the current'which can he cnlculated from a knowledge of tho potential gradient rtncl t.he coii- ductivity of the air just abovr: the surface. Measure- ments have been made of bot-h these f actois b Gerdim ao
that in 31 heights of the atmosphere the potential radient
words the same vertical current is resent throughout the
carrying measuring instruments. Now, as the same vertical current is resent in 6,000
meters a3 on the
with the positive electricity in t i e air. This leads to the important conclusion that t.lie negat.ive chn.rge on the surface and the positive charge in the air are not comple- mentary in the sense that one has been estracted from the other. The relationship between these two charges
of the emanation there can be no hig g ionization, for the
flon through the c K annels and cracks due to ti falling
has been suggested to account 9 or the separation of the
I have already state 2 that the loss from the surface is
to a hei ht of 6,000 feet (1,82S.S nieteisj, an K it is found
multiplied by the conductivity is a consttint. P n other
atmosphere up to the greatest heig \ t reached by balloons
ound, this ineans t l a t none of the negative charge w T ich left the round has combined
'
P
tain time the accumulated charge on the earth has risen until i t has set up, in the atmosphere, at field sufficiently reat tozany away the char e as rapidly as it is received.
earth receives its constant sup ly of elec.tricity. When
us1 quantities of electricity passing upward in ~1 1 parts
2 the atmosphere. If the coaductivity of the air were the same throughout the atmosphere it would need the same force to drive the same current through all layers. But the air does not have the same conductivit through-
!Ius is a final steady state w Fi; ich continues !omlong as the
this state has been reached t g roughout there must be
out, the conductivity increases with height an t f therefore
10 Gerdien in Nachr. d. k. Geaelhh. d. Wis#M. zu Gettlngen, 1805. Heft 5.
the force necessary to drive the current decreases with hei ht; in other words, tho potential grsdient decreases
ductivity so that a constant current can flow, can o 11y be effected by the accumulation of B volume charge of elec- tricity.
The following is a statenlent of the process by which
the steady state is reached. The charge on the earth itself incrertses until it produces such a field in the air
immediately above the ground that the electricity is con- ducted from the surface at the rate a t which it is received. Now, in another layer some distance above the ground the conductivity is greater, hence if the fiald due to the charge on the earth extended so high it. would produce too large a current through the layer; in other words., the lnyer would lose electricity faster than i t received it from below. The consequence would be that iincoinpensated positive electricity would appear in the layer. This would counterxt some of tbe field due to the negative charge on the round and the field i n the layer would decrease.
so induced, woulcl be cut dowx until a stead state was reached in which the current from the grounJwould just be condwted through it and no more. This process would go on throughout the at#mosphere; in every layer sufficient free positive charge would a pear to reduce the
through it.
Thus the two independent varishles are, a, the rate at which t,he earth receives its rharge and, b, the ronduc- tivity of the atmosphere. Given these the potential gradient and the volume charge adjust themselves until the same quantity of electricity passes through every layer of the atmosphere. But these are exactly the conditions which we find in the earth's atmos here; the charge on the earth, the volume charge in t l e air and tha potentia! gradient are so adjusted that the esisting conductivity is 'ust sufficient
the atmosphere. We inust therefore conclude that the
prime cause of the electrical field in the atmosphere is that the earth receives n constant charge of ne ative
wit % height. This adjustment of the force to the con-
Fim f ly the field in the layer owing to the positive charge
field to the value required to coaduct t 1 e constant current
to carry a constznt vertical current throug i all lsyers of
electricity which must be conducted away as rapi 5 ly rn
it is received. We can go further and say that as observations have shown thaz the same currcht passes the 6,000-meters layer as leaves the ground, the supply of electricity to the around can not be obtained from the atmosphere below -
~
this height.
We now see whv all attemnts to solve the problem of the earth's ne at& charge hfconsideria proc&ses which
failure.
To solve our problem then we have to discover some means by which the eart,h may receive a constant supply of negative electricity without the corresponding positive charge being set free within the earth itself or the lower
are mnipletef wit,hin the lower atnlosp fa ere have led to
atmosplpre. This 15 a much more difficult problem and one which, it may be stated a t once, bafAes all known physical pro- cesses. When lon and serious attem ts have been made to
appears to me that it IS justifiable to draw on one's imagi- nation and to atate what kind of rocess robable or not,
would satisfy the conditions. 5 therigre pro ose to
solve a prob K em along recognize cp lines without result, it
describe two proceases which would explain the p i enom-
122 MONTHLY WEATHER REVIEW. MARCH, 1916
(a) The balls appear to esist inde endently of any large
and have also been observed to pass in and out o parallel telegra h wires.
(c) #hey appear to be associnted directly or indirectly with large quantities of energy, for they have been ob- served to explode with violence nnd have dso been seen to fuse the overhead wire of an electric traniw-ay. There have been, of course, many atteni ts to explain
most explanations the seat of the low is su posed to be
itself is of the nature of a brush discliar e ; this,..however,
til recently we had no knowledge of glowing gas except when associated with an existing electrical discharge or
a flame. Prof. Strutt 22 has, however, shown tlmt by means of an electrical discharge a mass of nitrogen can
be put into a state in which it continues to glow for some time after it has been removed from the field. I c m not help believing that the body of the ball liglitnin is some
of a lightning flash. Prof. Strutt has pointed out to m e
serious difficulties of this esplanation amd I can see others, but future work niay remove t,hein. In any case active
nitrogen is the nearest physical phenomenon to ball lightning yet produced in our laboratories. I feel that
this subject has not received the attentioii it deserves by experimental physicists, and esperinient's made t.o solve this problem might well lead to most important results.
electrical intensity, for the have R een observed within closed rooms where large e T ectrical fields are im ossible,
P
this phenomenon with an entire want o P suc,ccss. In
associated with intense electrical P orcn so t RR t the glow
is almost certainly not the case with ha K 1 lightning. IJn-
gas made to glow in this way by the intense c f ischarge
6. NATURE AND ORIGIN OF THE AURORA.
I now come to the last and probably the niost inter- esting of the probleiiis which I propose to discuss. Many years ago Rirkeland 23 suggested that the aurora is due to electrified parbicles shot off from tlie sun, and
as was only natural, he considered these to be of the same nature as the negative obctrons with which we have become so familiar in recent work on cat,hode rays and radioactive suhsttincc?s. Rirkeland su gested that
but those which pass near to the earth got entangled in the earth's magnetic field and travel alon t'he niagnt*t,ic
duce the offects of the aurora when t,hey st,rike the uppcr regions of the atmosphere.
iiiatheniatical coniputations niade by St,ijrnier.?' It was realized throughout t,hat both the esperiiiients and computations left open the questioii of tho sign of the charged particles, but the negative electrons were considered to be the most likely for many reasons, not the least being that glowing gases such as exist a t the surface of the sun are known to emit a copious stream of negative electrons. In 1913, however, Veuard 3 pointed out that, judging from the characker of tKe aurora rays, it is more likely that the aurora is produced by the impact of the chnrgod positive articles correspondin with the a rays of radio-
out a very strong case.
these electrons are discharged hy t,he sun in a 9; 1 directions,
lines of force toward the magnetic poles w 91 iere they pro-
by remarkable laboratorj- experiments mid by e freed a orat.e This theory was su
active su \ stances, and it must s ,e adiiiitted that he mado
Strutt BakerIan Lecture i?r Proc. Ro sop 1911 85A: 219, and in numerous sub- mqumt pipera In the Proceedings of the d&l'So&y. Bhkelaud in ExWtIon Norm ienue de 1899-1900. Stllrmer sn CMatiama Videnakafxsehkoda Skr., mat.-nat. l i l . N:o a, 1901.
ab V-d tn Phil. IMg., Londa, 1011, 111.
During the last few years Prof. Carl Stormer has worked out with great patience and success a method of determining the exact hei ht and position in space of the aurora rays. His metho d consists in photographing the aurora simultaneously from two stations a t a consider- able distance apart, which in practice is 37.5 kilometers. Then by comparing the a parent position of any marked to the stars on the two plates he is able to calculate the coordinates of that part of the aurora. He has recently published a preliminary report 2a which is of surprising interest. He shows how the aurora rays end very suddenly between 90 and 100 kilometers above the earth's surface, but the result which is of the most interest is that in one marked case he has been able to determine the sign of the electricity in the rays. By calculating the exact position of the aurora ray at a certain time and comparing this with the simult& neous magnetic disturbance at the earth's surface he found that the ra must consist of positive electrons. I t
worked out one case, and lie himself asks that the rem t should be taken as provisional until lie has worked out further esamples. At the same time, as this observation fits in so well with the considerations advanced by Vegard, I think we are 'ustified in considering the consequences of this re-
Vegard shows that the a rays which cause the aurora have the characteristics of the rays emitted by radio- active bodies, and his arguments point strongly to their origin being actually radioactive substances in the sun. I t is a fascinating occupation to consider what may be the harvest of this discovery if future work should con-
firm it, and I can not refrain from mentioning some of the thought-s to which it already gives rise. In the first place, if the sun contains radioactive mat,ter to such a large extent as to give the copious radiation necessar? to produce the aurora so far awa as the earth,
his energy must be affected. What must be the electrical field in the region sur-. rounding the sun clue to the constant loss of so much electricity? And what becomes of the matter and elec- tricity distributed in this way to regions far beyond the earth's orbit 1
Then, again, as the a! radiation from different radio- active substances has different characteristics it may be possible to determine from tlie nature of the aurora what are the radioactive substances actually prevent in the sun. If the (Y radiation from the sun produces the aurora, what becomes of the /3 and y radiations which the same
radioactive substances must emit? These may be the cause of the high ionizstion of the upper atmosphere which wireless telegraphy and Schuster's theory of the daily variation of tlie earth's magnetic field have led us to expect. The depth to wliicli the radiation penetrates into our atmosphere ma.y give 11s information as to the density, and therefore temperature, of the u er atmos here m
earth's surface. These are only a few of the vistas opened up by this great discovery, and each one of you can doubtless sug- It is obvious in any case that another great
geld of cosmical discovery has opened up, but unfor- tunately it is one in which we in India are unable to par- ticipate. We can at least wish our confrgres in polar regions all success in their work.
feature of the aurora reative P
T
must be stated, i! owever, tdhat Stormer has so far 0111
mar k able result.
all theories as to the nature of the sui1 an B the supply of
regions far higher than we can possi Fl y reach P rom the
est others.
StLmer in' Terr. magnet. and atmosph. electrfdty, lQl5,10: 1.