E.R.A. Report 5275
M.A., D.Sc. (Edin.) F.I.E.E., F.Inst.P. F.R.A.S.
(Research Physicist in The Electrical Research Association from 1924 to 1967)
The Electrical Research Association
The object is to show that all cosmic atmospheric phenomena can be explained as deriving from electrical discharges, resulting from the breakdown of electric fields generated by the asymmetrical impacts between dust particles, such as are effective in terrestrial electrical sand and dust storms and in thunderstorms. These electrical discharges form, for example, the solar photosphere at 6,000°K, superposed on an atmospheric background temperature of less than 4,000°K at which solids can and do form. Isolated discharges form the solar prominences and solar flares. The electrical discharge theory of the latter led to the prediction (1959) that they must emit X-rays before these were observed by the first U.S.N. satellite observations in 1960 and observations of the transverse magnetic fields surrounding two flares in 1966 by Severny have confirmed that the flares are, in fact, electrical discharges. Despite over 50 years of observations of longitudinal magnetic fields in the umbra of sunspots and of gas velocities limited to around 2 km per second in the Evershed effect, Severny confirmed that the former are actually transverse and Bumba confirmed that the latter reach 8 km per second, each in accordance with the theory's predictions.
The theory led to the view that, much of what has been regarded as due to stellar rotation is not due to the rotation of the star, as is now generally agreed, but to rotation of balls of gas emitted from discharges and this in turn led to a new theory of ball lightning which seems to explain all the observations. It also explains the photographs of barred spiral nebulae and of flashovers of insulator strings, in which the same escape of gas occurs at the sudden bends in the discharges at the ends of the bars and at the bends in the crinkled discharges over insulator strings in the laboratory.
The theory explains the origin of the cosmic atmospheric magnetic fields and relativistic electrons which have to be postulated by all other theories of radio galaxies; also of the gas jets which explain solar magnetic storms and the movements observed in the planetary nebulae, novae, and many extra-galactic nebulae. These discharge-generated gas jets derive from the same claracteristic of electrical discharges which renders arc welding possible.
The theory led to the value of 1060 ergs for the total energy liberated by a radio galaxy before this value was confirmed by Heeschen's study of all the available data. It explains the separation of the stars in a galaxy into two Stellar Populations and the observed differences in their average atomic constitutions and the two-armed structure of both stellar (planetary) and extra-galactic nebulae.
The introduction of atmospheric electric fields would appear to do for cosmic atmospheric astrophysics what the introduction of gravitation did for dynamical astronomy.(1) Introduction
A few years ago in a lecture to the British Association(1) Bondi emphasized the desirability that a theory should "live dangerously" by making predictions which could be verified by observation or experiment. The study of terrestrial lightning may be said to have been initiated by one of the tersest examples of this procedure when Benjamin Franklin ended his entry in his "minutes" on the lightning flash with "Let the experiment be made", and the Philadelphia experiment, as it was called in Europe, was successfully made, first by d'Alibard in France in 1752 and later by Franklin himself.
Unfortunately in proposing a step of even greater ratio in the extension of the field of electrical discharges in gases to cosmic atmospheres the writer has seldom been able to put his predictions in such a direct relation to possible experiments. What he has been able to do frequently is to suggest that the available observations, or even, and often, those not already available, should be studied from the point of view of their having originated in electrical discharges. Few theories can have lived more dangerously in this way and survived than has the electric field and discharge theory of cosmic atmospheric phenomena and universal evolution. It predicted(2.1), for example, the existence of two-armed nebulae on a stellar, as well as a galactic scale, when no one seemed able to confirm the existence of such a nebular structure -- not even theose who had worked for many years on planetary nebulae. It predictcd(2.2) the existence of temperatures of over 100,000,000°K in electrica1 discharges in the solar atmosphere -- soon verified(2.3) by U.S.N. satellite observations; of velocities of up to 8 km/sec in the Evershed Effect(2.4) despite over 50 years of observations limited to 1 to 2 km/s. Perhaps; the most remarkable success of all was the deduction that the magnetic fields in the umbra of sunspots must be transverse(2.5) despite the fifty years of observations at Mount Wilson Observatory and elsewhere purporting to show that they are not transverse but vertical. All the evidence would thus seem to indicate that these last predictions are incorrect in view of the accounts of sunspots and the Evershed Effect given in every text book on astronomy during the last fifty years or so. But possibly even more striking was the prediction of the existence and approximate duration(2.6) of what have since been discovered and called quasars, at least, four years before their discovery.
These and. many other successful predictions of the theory are described in the present account. As each contributes to the significance of its fellows, it seemed desirable to collect them together with those not already discussed. This has been done under the heads of solar, stellar and galactic phenomena. It might have been more logical to have collected them under such headings as electric field building, discharge characteristics, discharge-generated gas jets, thermal and non-thermal cosmic energy sources, etc., but the present arrangement will probably appeal more to those who specialize in solar, stellar or galactic phenomena.(2) The Sun
The theory may be said to have started with a successful prediction during Sydney Chapman's Kelvin lecture(3.1) on the sun to the Institution of Electrical Engineers on 8th May, 1941, when he referred to a solar prominence which had reached a height of a million miles in an hour. It seemed to the writer that this could only mean that the phenomenon must be a solar lightning flash and that, therefore, a million miles an hour must be equivalent to 3 x 107 cm/sec, the velocity of propagation of the lightning leader-stroke since the velocity of propagation of breakdown in a gas should be independent of the density. A little mental arithmetic in the darkness corroborated on a scrap of paper when the lights went up, verified that this first prediction was approximately correct and the electrical discharge theory was launched. It soon appeared(2.7) that this prominence could not be an isolated electrical discharge but that all the solar "surface" phenomena must be electrical; that the granulations of the photosphere, the spicules of the chromosphere and the rays of the corona, form a hierarchy of electrical discharges at decreasing gas densities and increasing discharge temperatures.
This general conclusion had been the more easily reached as the writer was at the time working on the lightning discharge and had shown(2.8,2.9) that the electric field required to cause a lightning flash is only about 1% of that which C.T.R. Wilson had suggested(4.1) as necessary and which earlier theories had taken for granted (5,6). The' onset of instability and breakdown was shown to be marked by the transition in the field-maintained corona or glow discharge of St. Elmo's fire having a positive characteristic to the thermally ionized channel of the arc discharge, possessing a negative characteristic. So far as the writer is aware all subsequent criticisms of that theory, and of the corollary that lateral corona currents must flow from the leader-stroke channel, have since been answered(2.6,2.10), and Pierce later found that the theory enabled all the observational data obtained at Cambridge to be correlated "with no major inconsistencies"(7).
It will be evident that the theory at once explained one of the outstanding problems presented by the study of the solar surface, namely its "fibrille" or "granulated" structure, the individual granules being individual discharge channels. It is supported by the very careful study of the formation and life of individual granules made by Bray and Loughhead who write that "in general a granule develops from a vague patch of diffuse bright material, which originate's in a hitherto dark area."(8.1). (The underlining is the present writer's).
The first publication in a scientific journal appeared in "The Observatory" in April, 1946 (Ref. 2.10a) and summarized those characteristics of solar prominences which were at once seen to conform with the expectations of the discharge theory, including their extreme thinness, which had often been commented on, the fact that some are initiated as much as 150,000 km above the "surface" of the sun, and that these downward developing prominences on occasion show tho leader-return stroke mechanism of the lightning flash, and that one prominence can exert an influence on others over 250,000 km away. They also on occasion simulate lightning flashes in that successive strokes follow one anothor along the same discharge channel. As Ellison had written "one gets the striking impression that the trajectory is a conducting path, quite distinct from the prominence material which is constrained to move along it".
It was also pointed out that "the very fact that such streamers can be initiated as much as 150,000 km above the surface of the sun itself support the writer's conclusion(2.7), reached on more general grounds, that the sun and stars must be surrounded by extensive tenuous atmospheres, and has been so interpreted by Pettit himself" (Ap.J., 98, 1943, p.6) who first, recorded this type of corona prominence.
(2.2) Solar Atmospheric Electrical Fields
The above quotation from Bray and Loughhead leads naturally to the question of the setting up of the electric fields which lead to the development of bright granules in these dark, low temperature areas. Chapman has written(3.2) recently "Bruce agrees that the sun offers his ideas perhaps their greatest challenge, because of the very high electrical conductivity of the solar material at all levels".
On the contrary, the solar atmosphere offers one of the best checks of the theory and confirms a prediction which has been implicit in the theory since 1955, when a possible explanation of these atmospheric electric fields was put forward(2.11). Till then the necessary electric fields were merely, like so many of their magnetic counterparts, a hypothesis. Unlike these latter hypotheses, however, the electric field was the one hypothesis which made the whole universe kin, so to speak, and obviated the need for the many other hypotheses including those concerning the magnetic fields themselves.
Since 1955 the discharge theory inevitably led to the tacit prediction that a surface of discontinuity must exist in any stellar atmosphere when that temperature is reached at which solids begin to form, i.e., at temperatures of around 4,000°K. At these temperatures hydrogen is far below its dissociation temperature and still further below the temperature required for appreciable thermal ionization. The theory itself supplies the convection currents even if they were not there in the absence of the electrical discharges, so all the factors are available to lead to electric charge separation and field building, thanks to the observations which can be made in the absence of the discharges, i.e., in sunspots. In these, when we can see further down into the solar atmosphere in the temporary absence of the photospheric arcs, this thermal background is exposed to view and is at the temperature predicted by the theory, 3,700°K or 3,800°K (Ref. 8.2), more than 2,000°K less than that of the discharges. It is only when this last temperature, about 6,000°K is reached, i.e., that required for thermal ionization of the gas, that the setting up of electric fields becomes impossible.
It may be emphasized too that de Jager(9.1) has given this same temperature, around 4,000°K, as the background temperature of the chromospheric discharges.
It is thus fortunate that we are able to see the details of the sun's atmospheric structure in sunspots, and verify that it conforms to the picture which the discharge theory had led us to expect; that is, a general background atmospheric temperature of around 4,000°K in which electric fields can be built up by asymmetrical impacts between solid particles, just as occurs in terrestrial sand and dust storms and in the ejectamenta above volcanoes. When these fields reach the values required for electrical breakdown, then the latter results in those temperatures being reached which are necessary for thermal ionization of the gas, the 6,000°K of the photospheric arcs. When the field in its neighbourhood is neutralized, then the arc is extinguished, the gas cools, and the process is repeated.
Bray and Loughhead in the passage already referred to(8.2) note that "the dissolution of a granule appears to occur by the reverse process, although occasionally a granule loses its identity by coalescing with another granule", which is not surprising since these parallel currents will tend to attract one another, just as has been observed in solar prominences.
(2.3) Solar Flares
(2.3.1) Crochets and Balmer Line Widths
Crochets are those sudden variations in the earth's magnetic field caused by the occurrence of solar flares and it seemed to the writer that they might be the atmospherics radiated by the solar lightning flash, especially as some of those delineated by Newton(10) had a wave form similar to those of terrestrial atmospherics(2.12). As the writer had just then developed a method of calculating the latter based on his theory of the leader-return stroke mechanism in which lateral corona currents flow from the discharge channel during both these phases of the lightening flash, the same calculation could be applied to calculate the atmospheric radiated from a solar flare(2.13) The values of the current and current density obtained, 1014 amperes and 10-3 to 10-5 A/crn2 respectively, seemed quite reasonable, as did the total rate of energy radiation from the discharge, 10-5 of the sun's total output. There seemed to be one insuperable difficulty, the theory led to the conclusion that the magnetic fields in the discharge channel itself must reach 104 to 105 gauss, and no one had previously mentioned the existence of such fields though fields of several thousand gauss had been observed in sunspots. However, it was verified that magnetic fields of this order will tend to Ha line widths between 0.8 and 8A, and a search of the literature brought to light Ellison's(11) observed range of 1 to 16 A, in good agreement with the theoretical values.
Until quite recently there remained one other apparent difficulty. The maximum magnetic fields associated with solar flares observed with Babcock's solar magnetometer(12) only reached 1000 or 2000 gauss. The writer has shown(2.14) that this is due to an instrumental limitation, so to speak, since the instrument uses the lines of neutral iron. The spectra of solar flares contain lines of levels of excitation up to Si XII, i e., up to levels of the order of 500 eV, to say nothing of high energy X-rays(9.2). Applying the "ionization potential thermometer" to which the writer drew attention in an earlier note(2.15), this means that the discharge temperature must reach values of the order of 500,000°K. As this temperature is higher than any mentioned in his earlier notes, it may be pointed out that the "readings;" on this thermometer have since been checked up to discharge temperatures of around 3,000,000°K in laboratory electrical discharges(2.16), and they do not depend on the nature of the heating process of the gas(13).
Theory and observation thus lead to a conception of a solar flare as possessing an axial temperature of the order of around 500,000°K to 1,000,000°K, a diameter of the order of 50,000 km, and a magnetic field of 104 to 105 gauss at its surface, i.e., at, a radius of 103 to 104 km. In relation to this description the pertinent characteristics of the solar magnetometer are, first, that it mainly uses the Zeeman effect in the lines g5250 and g4486 of the neutral iron atom, and second that it averages the field oven distances of the order of 105km. At distances of this order the magnetic field will have fallen to the values of 103 gauss recorded by the magnetometer and it is doubtful if there will be any neutral iron atoms to record a Zeeman effect much within this distance. So far as they go, therefore, the recordings on the solar magnetometer would appear to support the fields deduced from the discharge theory of solar flares.
The theory as developed in 1949 will be seen to lead to two obvious predictions: (1) the atmospheric radiated by the discharge and observed at the earth will be greater if the flare occurs at the limb and not at the centre of the sun; (2) the atmospheric will be greater the greater the current in the discharge, and hence the greater the magnetic field and the width of Ha in the flare's spectrum. In confirmation of these two deductions from the discharge theory Smith and Smith write(9.3) in the discussion of Dodson and Hedeman's observations: (1) "Smaller flares near the limb are more likely to produce crochets than are small flares near the centre of the disk"; (2) "Crotchet-associated flares appear to have a greater average maximum Ha line width than do other flares". The discharge theory would thus appear to answer the "challenge to interpretation" which Smith and Smith write on the same page is presented by "this peculiar distribution of the crochet flares".
(2.3.2) Flare Energy:
Another difficulty which the same 1949 theory clears up(2.13) is that found by Smith and Smith in their discussion(9.4) of large energy density of solar flares, 0.5 to 5.0 x 103 erg/cm3, which they contrast with the 1 to 10 erg/cm3 of the undisturbed chromosphere. However, just as in a lightning flash much of the pre-discharge field energy is liberated in the discharge channel, and in a laboratory spark discharge much of the energy stored in the apparatus is released in the discharge channel, so the energy of the pre-discharge field is similarly concentrated in the flare.
The figures obtained in the 1949 note(2.13) were, for the current, 1014 A, electric field, 8 x 10-2 V/cm, and the flare radius was taken as 103 to 104 km. Comparison with Ellison's observed line-widths indicated that either the current was greater than 1014 A or the radius was less than 103 km. If then we take the values given we may expect a lower limit for the energy density. The value so obtained is about 2 x 103 erg/cm3 which again agrees well with these subsequent observations referred to by Smith and Smith.
(2.3.3) Pressure in Flares:
In 1937 Bellaschi(14) showed that the axial pressure in an electrical discharge carrying a current I with a current density i would ultimately be increased by a factor10-8 Ii. With the values of current and current density in a solar flare obtained in 1949 this led to the prediction that the pressure in the axial regions of a flare should ultimately be increased by a factor of between 10 and 1,000 according as the current density is 10-3 or 10-5 A/cm2. In 1961 Jefferies and Orrall(15) confirmed this deduction from the discharge theory when they found that "the gas pressure must have been ten times higher at the centre of the flare than 20,000 km out", since the value they thus quote for the radius of the flare corresponds to that which gave the smaller current density in the 1949 note on solar flares already referred to(2.13).
(2.3.4) Gas Jets and Magnetic Storms:
One of the great advantages of the electrical discharge theory has been referred to in the previous section. It has built into it the most powerful known atmospheric aggregative force(2.17). When it operates as it does in all electrical discharges from the welding arc to those which form the arms of the galaxies, gravitation cease's to have any significant effect. This leads to a second great advantage possessed by the theory(2.18), for this aggregative force varies with the current and the current density in the discharge which both vary along these discharge channels. Beyond some point both will decrease radially outward from the sun and will, therefore, result in a pressure gradient and hence a flow of gas along the flare(2.19). It is this plasma jet which is responsible for the main phase of the magnetic storms which are observed to begin suddenly at the earth about a day after the commencement of the solar flare(2.20). This represents a mean velocity of about 1,000 km/sec and H atoms have actually been observed entering the earth's upper atmosphere at velocities of up to 3,500 km/sec (Ref. 16), The velocities of these discharge-generated jets can often be used to determine the temperature of the gas(2.21), and it was pointed out(2.22) that these velocities indicate that the temperature of the solar flare discharge reaches values of the order of 1,000,000°K. This prediction had not long to wait for confirmation by satellite observations(2.23).
(2.3.5) Thermal and Non-Thermal Sources:
Recently(2.24) the writer developed a suggestion he put forward many years earlier(2.25) to account for non-thermal cosmic ray and other sources. It in turn was the development of a still earlier suggestion of C.T.R. Wilson(4.2) that the soft cosmic rays might be runaway electrons accelerated in the electric fields of thunderstorms. In 1941 the writer had shown that the over-all electric fields postulated by Wilson do not exist in thunderstorms but in 1952 he showed that the mechanism might be operative in the field-concentrations built up during the propagation of these cosmic leader strokes. Since the temperature of the discharge in solar flares reaches more than 108°K, the velocity of the H atoms reaches about 2 x 108 cm/sec and the velocity of the electrons reaches more than 40 times this or about 8 x 109 cm/sec. If the solar electric field is positive relative to the gravitational field, like the field in a thundercloud, then some of these high speed electrons can be pulled out in the field concentration ahead of the plasma jet and further accelerated. They will then account for the soft cosmic rays which arrive at the earth within half-an-hour of the start of a flare, indicating an average velocity from sun to earth of around 1010 cm/sec and an acceleration in the field by a factor of only about, 1.25.
Further support for this suggestion is afforded by the results of the study of the corresponding characteristic of galactic discharges in Section(4.5).
(2.3.6) Forbush Decreases:
Coincident with the arrival at the earth of the discharge and plasma jet some 24 hours after these cosmic rays, the earth will become enveloped in the discharge's magnetic field. This will have two effects: it will prevent soft cosmic rays from reaching the earth, and so account for the observed Forbush decreases in cosmic ray activity observed to accompany rnagnetic storm(17). it, will also explain how the charged H atoms of the solar stream reach much lower levels in the earth's field and atmosphere than they otherwise would, without having to make the quite drastic assumption of Akasofu's theory(18) of magnetic storms that atoms of gas can travel at these velocities and yet be unionized.
(2.4.1) Explanation and Structure
Whereas most astrophysical theories were and probably still are embarrassed by the observation of the low temperatures of sunspot's in which we are able to see further down into the sun's atmosphere, this relatively low gas temperature naturally confirmed one of the discharge theories earliest tacit predictions, that, the temperature of the atmosphere surrounding the photospheric arcs must, be much lower than that of the discharges themselves.
It was suggested in the original summary of these ideas(2.7) that the cause of the extinction of the photospheric arcs is the occurrence of a much larger discharge or facula which neutralized the electric field in their immediate neighbourhood. It was then emphasized in support of this view that, "sunspots are always accompanied by faculae, and are in fact preceded by them", as had been then determined at Greenwich and as is generally agreed(19.1).
The existence of the penumbra too was readily accounted for and followed from the attraction on one another of parallel currents. The current decreases outwards in these atmospheric discharges since the current passing any point has to maintain the lateral corona currents beyond that point. It follows that the the discharges near the "surface" of the photosphrere will be pulled outwards from the centre of the spot and downwards so that a vertical section through a spot will have the form shown in Fig.1, in which the width of the individual arc channels represents diagrammatically the magnitude of the current.
Fig.1: Diagram of vertical section through a sunspot
(2.4.2) The Evershed Effect:
Each of the arc channels in Fig.1 will have associated with it an axial jet of hot gas its velocity depending on the arc temperature. It followed(2.26, 2.27) therefore that there should be jets of hot gas moving along these bent discharge paths at the velocity of sound in ionized atomic hydrogen at 6,000°K or 8 km/sec. Evershed(20.1) had observed the flow of gas, which the theory leads us to expect, in 1909, but its velocity as observed by him and his successors in the following fifty years and more had been 1 to 2 km/sec. The writer had enquired at Oxford, where these velocities had been studied for ten years or more, whether they had found evidence of the higher velocities which the theory predicted. They could not but the higher velocities up to 8 km/sec have since been observed by Bumba, as was announced by Severny(21.1) at an I.A.U. Symposium on cosmic gas dynamics in 1961.
(2.4.3) Sunspot Magnetic Fields
One of the outstanding difficulties faced by the discharge theory during its whole life, has been the divergence between it and. observations on the magnetic fields of sunspots during the last 50 years or so at Mount Wilson and other observatories and the resulting "classical picture" of sunspot magnetic fields illustrated in every text book on astronomy. For it will be obvious from Fig.1 that the magnctie fields, being at right angles to the photospheric arcs, must be transverse through-out the umbra, and that only over the surrounding rim, or penumbra, of the spot should there be a vertical or longitudinal component. Much effort had been expended by the writer during the last twenty-five years in an endeavour to make the theoretical transverse fields longitudinal but to no avail, and the effort turns out to have been entirely wasted, as observations made by Evershed(20.2) in 1941 would have demonstrated had the writer but known of their existence.
It was gratifying and surprising(2.28) to read in Bray and Loughhead's recent book on "sunspots" at the end of a long section describing the "classical picture", "that the observations of Evershed and Severny(21.2) thus suggest a field of configuration very different from the classical picture. On the new view a predominantly transverse field in the umbra is surrounded by a large vertical field in the penumbra". Severny finds zero vertical field in some spots, in which the field is entirely transverse, and that vertical fields are observed only over the penumbra of spots, all exactly in accord with the predictions of the electrical discharge theory since its initiation in 1941.
(2.5) Solar Discharge Temperature
In all stellar atmospheres, that of the sun included, the discharges are usually propagated outwards, that is they are propagated down a density gradient. It was emphasized(2.17) that these long atmospheric discharges act 1ike "energy pumps".
We see the same effect in a lightning flash. The electrical energy is generated five to ten kilometres up in the thundercloud, but the leaderstroke short circuits the field between cloud and earth, and the largest current flows just outside the earth's surface. Just so, in these stellar atmospheric discharges the leaderstroke short circuits the electric field so that energy generated lower down in regions of higher density in the star's atmosphere is liberated further out in regions of lower density and longer mean free path. It is, therefore, to be expected that the discharge temperature will increase as the discharge is propagated outwards. The 6,000°K of the photospheric arcs becomes the 10,000°K to 20,000°K of the chromospheric glow discharges and this in turn becomes the 1,000,000°K or 2,000,000°K of the discharges of the corona. It was, therefore, not altogether surprising when the conclusion already mentioned(2.22) had to he drawn that the discharges associated with solar flares actually reach temperatures of over 100,000,000°K.
A similar rise in temperature is observed when discharges are propagated outwards in the denser atmospheres of the combination-spectra stars but having a lower density gradient when the 5000°K to 10,000°K of the initial discharges increases after a period of the order of 100 or 200 days to temperatures of hundreds of thousands of degrees and lines of Fe X and Fe XIV appear in their spectra(2.11).
(3) the Stars
When the general conclusion was reached in 1941 that all cosmic, atmospheric phenomena derive from electrical discharges, a conclusion at once followed which was disturbing to one who knew little about the results of astronomical research. If the outbursts in the long-period variable stars which last for about a year and if the nova outbursts which can last for ten years or more, are the result of electrical discharges, then, since the velocity of propagation of an electrical discharge is independent of the gas density, it followed that stars must exist with atmospheres which can contain such discharges. The velocity of propagation of the lightning leader-stroke is about 3 x 107 cm/sec, so the stellar atmospheres required must be of the order of 1014 or 1015 cm in depth, i.e., of the order of our whole planetary system. This seemed likely to remain a considerable theoretical stumbling block until the writer read of the existence of such stars in a supplementary note at the end of Russell, Dugan and Stewart's "Astronomy"(22) Fortunately such stars as e and z Aurigae, which are of this type, are members of binary pairs revolving in the line of sight, the other star being a small hot star. Some considerable time before the latter star goes into complete eclipse behind the larger star absorption lines begin to appear in its spectrum due to the passage of its light through the extended atmosphere of the other star, demonstrating the existence of just such extended atmospheres as the theory had predicted. An apparent stumbling block thus became a correspondingly strong argument in the theory's support.
The "new" observation was not entirely unexpected for it had seemed most improbably that the Sun could go on for 109 year, emitting the jets which had been postulated to account for magnetic storms without giving rise to an extensive tenuous atmosphere which the discharge theory of magnetic storms also required. The explanation of these extended atmospheres is given in section (3.7).
(3.2) The Long-Period Variable Stars
(3.2.1) Electric Field Generation:
If, as was seen in Section (2.2), the conditions for the build-up of electric fields are met in the solar atmosphere, by so much the more are they met in the atmospheres of the long-period variable stars. Even their "surface" temperatures reach only 1,500° to 4,000°K and the light from these surfaces in passing through their outer atmospheres is subject to considerable absorption by such molecules as TiO and ZrO which dissociate at around 1,600°K and 2,500°K respectively. Furthermore, the Doppler displacements of the emission lines when they appear during the outbursts show that gas movements of up to about 10 km/sec occur in their atmospheres. The conditions for the generation of electric fields as discussed(2.11) in 1955 are, therefore, seen to be amply met in their atmospheres.
This conclusion is confirmed by the explanation offered to account for a large part of the variation of the stars' light. The addition of the bright emission lines in their spectra, which denote the occurrence of the discharges in their atmospheres and which lead to their detection as long-period variables, account for a part of their variation in brightness, but much of the variation is due to veiling of the stars at minimum by the increased dust in their atmospheres(23.1).
(3.2.3) Epoch of Appearance of Emission Lines:
Merrill writes(23.2) that "Displays of bright lines near minimum light are specially curious" and for this reason refers to them later as "the bizarre emission lines". On the contrary the emission lines appear in the spectra, of the long-period variable stars just when the theory predicts that they should(2.11, 2.29). As the general atmospheric temperature falls after a "thunderstorm season", then the amount of dust in the atmosphere will increase and with it the rate of electric field generation. In the absence of any other cause of energy liberation this will go on until the field builds up to the breakdown value. It follows that, the latter will occur with the resulting emission of the bright line spectrum somewhere around light minimum.
The periods of light variation in the long-period variable stars will depend mainly on the time required for the electric field to build up to the breakdown value. A rough comparison(2.11) with the time required for this process in thunderclouds, taken as 102 sec, taking into account the main factors involved, namely gas density and velocity, and the gravitational field, predicted that the times required would lie between 106 and 109 sec. The observed periods of these stars lie largely between 100 and 600 days or 107 to 108 sec.
(3.2.5) Duration of Bright Line Spectra
Similarly the theory predicts that the bright emission lines will be emitted for a time obtained by dividing the atmospheric dimensions, 1014 to 1015 cm, by the velocity of propagation of electrical discharges, 3 x 107 cm/sec or the order of 107 sec. As the bright lines are emitted for times of the order of half the stars' periods, the theory's prediction is again in good agreement with observations.
(3.2.6) Variations of Period
In putting forward the new theory of the lightning discharge(2.9) this writer showed that it accounted for the considerable variation observed in lightning currents. The occurrence of these long atmospheric electrical discharges depends on the coexistence of two factors: (1) an average field adequate to maintain the discharge when once it is initiated by the transition from a field-maintained corona type of discharge to a thermally ionized discharge column; (2) in this low average field there must exist a field-concentration adequate to initiate this transition. In New York, for example, this transition occurs readily at the top of the Empire State Building, which initiates upward leaderstrokes(24). In purely atmospheric discharges or discharges from cloud to earth it will be effected by elongatod volumes of space charge in the thundercloud. The smaller the field-concentration the longer will breakdown be delayed, and the greater will be the current in the discharge when it does occur. Just as this consideration accounts for the observed variation in lightning currents, so it leads us to expect(2.30) that the maximum brightness reached in these stellar thunderstorms and the period-duration will also vary and this is observed.
(3.2.7) Variation of Spectrum with Epoch
From the general consideration that the time required for the electric field to build up to breakdown will be less the greater the density and the greater the gravitational force, breakdown will start low down in the star's atmosphere, and the discharges will be propagated out towards its periphery(2.11). When the emission lines first appear, therefore, the light; will be subject to maximum absorption, and the effects of this atmospheric absorption will decrease as the period progresses, on the present view of these outbursts. This is in fact what is observed to happen(23.3). Groups of lines, such as the Balmer series of hydrogen, and smaller groups, or multiplets, in the iron spectrum, for example, have definite intensity ratios in laboratory spectra. When the bright lines first appear around the time of minimum light the relationships between these line intensities are found to be considerably modified by the differential absorption to which they are subjected on the way out by the molecules of metal oxides, etc., in the atmosphere. This mutilation of the customary relationships gradually decreases as the outburst continues, until, towards its end, their ratios approach those observed in the laboratory.
(3.2.8) Change of Spectral Type:
At any point in the evolutionary progress of one of these stars the occurrence of the discharges in its atmosphere followed by a period of cooling and field-building will cause its general mean atmospheric temperature to vary between two approximate limits. If the whole atmosphere is gradually cooling, for example, a condition in which carbon particles (freezing point ~3,500°K) are formed at minimum and vaporized at maximum brightness, will be followed by a phase in which, say, zirconium oxide (freezing point ~2,500°K) plays a corresponding role. When the mean temperature falls by another 1,000°K or so, ZrO will be replaced by TiO. There are in fact three types of these long-period variable stars, and they are distinguished by the appearance in their spectra of the absorption bands of these three molecules. Stars of the two Classes R and N show the bands of C2 and CN; Class S stars show the bands of ZrO; and Class M those of TiO. These differences have usually been attributed to differences in the chemical constitution of the stars' atmospheres, but for the above reasons the writer suggested that they may instead be due to differences in the physical conditions(2.29). This suggestion, or prediction, is also subject to observational check. On the earlier chemical explanation, it would be quite impossible for one star to change type during an outburst, whereas on the discharge theory's physical explanation it would be quite possible for all the molecules of TiO, say, to be dissociated during an unduly high maximum, so that its bands would disappear from the star's spectrum, and at the same time sufficient particles of the normally solid ZrO could be vaporized, so that its bands would replace those of TiO. Fortunately such stars do exist. c Cygni is one whose spectral class occasionally changes from M (bands of TiO) to S (bands of ZrO) as the result of specially great outbursts(25).
It would thus appear that these stars are not necessarily in conflict with the general uniformity of chemical constitution of matter observed throughout the universe, as they were generally considered to be, nor do they necessarily indicate a trifurcation of the stellar evolutionary sequence in the way they are generally regarded as doing.
(3.2.9) Gas Velocities
The gas velocities observed in these stars afforded the first extra-terrestrial application of the gas-velocity thermometer(2.18). It had been shown(2.6) that the jets of gas generated when a lightning flash hits the earth account for metal atoms being carried half to five meters up the channel during the times available and thus explained Isräel and Würm's observation of metal lines in lightning spectra, up to heights of about two metres(26).
The appearance of emission lines of hydrogen and ionized metals in the spectra of the long-period variables together with the ionization potential thermometer indicated discharge temperatures of 5,000°K to 10,000°K, so the theory suggested(2.18) that the gas velocities should lie between the velocities of sound in ionized atomic hydrogen at these two temperatures, or between the values of 8.5 and 12 km/sec. No satisfactory explanation for this outward flow of gas had previously been offered and considering the whole range of observed gas velocities up to thousands of km/sec this was a very narrow theoretical target, yet it contained the two mean values of 11 km/sec, which Merrill(23.4) obtained from a group of 72 long-period variables, and 9 km/sec, which his colleague at Mount Wilson, Joy(27), obtained from a group of closely associated irregular variables.
(3.2.10) Gas Velocities in AX Persei
Though AX Persei is not a long-period variable but a "combination-spectra" star it is sufficiently close to the former for the gas velocity of 110 km/sec observed in it by Merrill(23.5) to come as an unpleasant shock. Since the velocity of sound only increases as the square root of the temperature, this increase by a factor 10 in the velocity implied that the discharge temperatures in AX Persei must reach much higher values than they do in the long-period variables. This conclusion is confirmed by the observation(28) that the lines of highly ionized atoms, for example, 9 times ionised iron, Fe X, and 6 times ionized calcium, Ca. VII, appear in the spectra of AX Persei and associated stars such as CI Cygni.
(3.3) Cometary Nebulae
The existence of cometary nebulae such as that associated with the star R Monocerotis, Hubble's Variable Nebula show the result of the aggregative force of earlier discharges, probably of the nova type, in their atmosphere. In this region of increased density discharges are practically continuous and account for the variability of the light of the nebula. This view suggested an explanation for Beals' observations on the well-known star P Cygni(29).
(3.4) Gas Velocities in P Cygni and c Cygni:
The diagram (Fig.2b) of the conditions in the discharge in the star P Cygni are taken from an earlier paper(2.31)and the discharge channel is assumed to be towards us in the line of sight. The only point at which the gas should have the velocity of sound in ionized atomic hydrogen is in the throat of the expanding nozzle formed by the discharge channel, where the temperature is greatest. Beyond the throat the gas cools and accelerates, the flow becoming more and more supersonic. The highest level of excitation is 47 eV of the CIII and NIII ions, indicating a temperature of 47,000°K. The theoretical velocity is, therefore, 28 km/sec and this is precisely the velocity of the CIII ions observed by Beals. This was the first application of the ionisation potential thermometer to determine these discharge temperatures.
Fig.2 (a): Photograph of Hubble's Variable Nebula
|Fig. 2(b): Diagram of the discharge channel in the atmosphere of P Cygni.
In c Cygni the levels of excitation and the gas velocities observed by Merrill are both much lower(2.15), but show a similar increase in velocity with decrease in the level of excitation, and the lowest gas velocity is again close to the expected value appropriate to the highest level of excitation.
(3.5) Planetary Nebulae
(3.5.1) Two-Armed Structure
As mentioned in the introduction one of the most surprising conclusions to which the theory led was that two-armed nebula should exist on a stellar atmospheric scale quite analogous to their counterparts on a galactic scale and that photographs of them very probably existed in view of the existence of such photographs as that of Hubble's Variable Nebula. It was almost equally surprising that even those who had worked on planetary nebulae for many years were unable to refer the writer to them until Merrill wrote that the likeliest place to look for them was in a paper published 42 years earlier by Curtis at Lick Observatory(30), and there indeed they were found.
Fig.3: Rough sketch of expected form of two-armed planetary nebulae.
The prediction derived(2.1), first, from the explanation given for the spectrum of P Cygni, and its success naturally helps to confirm that explanation and, second, from the observation that stars like g Cassiopeiae exist in which there are not only emission lines displaced to the violet, but pairs of emission lines, one displaced to the violet, the other to the red. These could obviously be explained if double nebulae existed as in the diagram. If the two arms were in the line of sight then the two discharge-generated jets would give rise to the two emission components.
Merrill 's suggestion immediately led to the confirmation of the success of this prediction, as will be seen from the examples in Fig.4, taken from Curtis' paper.
Fig.4: Photographs of planetary nebulae showing pairs of discharge channels (Curtis: Lick Observatory).Note: The images of nebulae from this original article were too poor to reproduce, and have been replaced by alternative images from a variety of sources. The captions remain unchanged
(3.5.2) Gas Movements
A more detailed study (Ref. 2.32, Figs.4 and 6) of the gas movements in two of the ten nebulae which had entirely baffed Campbell and Moore(31) showed that they are at once separable into two jets, straight in NGC 2392, the type nebula of the group of ten and spiral, as its photograph suggests, in NGC 6543. Both spiral and barred-spiral discharge channels occur on a stellar as well as a galactic scale, but so far no irregular stellar nebulae have been photographed so far as the writer is aware. It may be that the small scale does not allow for the development of irregular nebulae, that there is always sufficient remnant angular momentum of the gas to lead to the development of a diametral rotational and, therefore, electrical plane.
(3.5.3) Feast's Data on Gas Velocities
The writer's friend, D.R. Barber, ex-Superintendent of the Norman Lockyer Observatory, very kindly drew his attention to the close agreement of the gas velocities observed by Feast(32) recently in a number of planetary nebulae, with those to be expected from the discharge theory. The mean level of excitation of the lines used by Feast is 36.2 eV, indicating a mean discharge temperature of around 36,200°K and, therefore, a mean gas velocity of about 23 km/sec. Tlie mean velocity observed by Feast is 22 ± 7 km/sec.
(3.5.4) Zanstra's Stellar Tempuratures
Zanstra(33) has also put forward a theory of planetary nebulae which is based on two fundamental hypotheses: (1) the nebula is a more or less uniform ball of gas; (2) it shines by fluorescent radiation deriving from the light of the central star.
Observations would seem to show that neither of these hypotheses is correct. On the discharge theory it is a two-armed structure shining by the light of the discharges still going on in it.
Two simple tests can decide between the two theories. In the first place photographs show that the nebulae are two-armed in accordance with the discharge theory's prediction. Those which give the impression of a star surrounded by a luminous ring are those in which we are looking along a discharge channel. They are those in which the gas jets are in the line of sight. Secondly, Zanstra's theory leads to a determination of the temperature of the central star from the intensity of the bright emission lines in its spectrum, so obviously the values so derived for any one star should all be the same. On the contrary on the discharge theory the result must depend on the temperature of the gas where the spectrum line is emitted and, therefore, the value obtained should be linearly correlated with the level of excitation of the line used in its determination. (See Fig. 11)
The writer, therefore, plotted the "stellar temperatures" given to illustrate Zanstra's theory in Ambartsumian's "Theoretical Astrophysics"(34) against the level of excitation of the lines, and it can be seen in Fig.5 that all the points fall reasonably close to the writer's theoretical line.
(3.6) Stellar Rotation
(3.6.1) Dependence on Temperature
Throughout this investigation, right from its initiation, there has been a close link up with laboratory and terrestrial electrical discharges. The origin of cosmic gas jets, for example, is the same as that of those which King(35) showed render arc welding possible. The ionization-potential-thermometer has been checked up to much higher temperatures in laboratory discharges than those to which it has been applied in cosmic electrical discharges(2.16) The discussion of what has been regarded for over 80 years as the effect of stellar rotation on line broadening is yet another good example of the inter-relation of phenomena of widely different magnitudes.
It seemed to the writer(2.33) that in all probability what is actually rotating at these very high speeds is not the star itself but huge volumes of gas escaping from the discharge channel. Again a simple test of this hypothesis is immediately available, as will be apparent from the diagram in Fig.2(b). If the gas escapes from the channel where lines of CIII are in emission its velocity would be 28 km/sec, Hd lines should show 189 km/sec and Ha lines 280 km/sec. D.R. Barber was, therefore, consulted as to whether he could supply examples of the widths of say the Balmer lines of hydrogen varying along the series. He replied immediately with the data obtained by Pottasch(36) from the spectrum of g Cassiopeiae given in the Table and it will be seen from the Table that another prediction was immediately successful.
Gas Velocities from Hydrogen Line Widths in the Spectrum of g Cassiopeiae.
(3.6.2) Stars of Types B and Be
Since on the electrical discharge theory the addition of the e in the denomination of stellar types indicates the existence of discharges in the atmosphere of that particular star, a comparison which immediately suggested itself was between the "rotational" effects observed in stars of types B and Be. Fig.6 (Slettebak(37)) again shows how well the expectations of the theory are fulfilled. The "rotational velocities" of Be stars are considerably higher than those of B stars.
(3.6.3) Flare Stars
Another type of star in which similar effects might be expected are the flare stars, the flares denoting the occurrence of huge discharges in their atmospheres. These stars also show much wider lines than are normal for their spectral type and this increased width has hitherto been attributed to high rotational velocities of the stars themselves(38.1).
(3.6.4) Current Theories
In "Stellar Atmospheres" there is a long chapter on stellar rotation by Huang and Struve the last page of which tends to throw the whole subject into the melting pot(38.2). In contrast to the writer's experience in seeking out evidence of the observed relationship between "rate of rotation" and gas temperature in support of the discharge theory's prediction, the authors are quite non-plussed by the finding of this evidence and write "since it is difficult to suppose that there is a correlation between stellar rotation and surface temperature or pressure, we again conclude that the observed line widths in super giant stars are not due to rotation". The suggestions they have to offer in the next two paragraphs, the last in their chapter, all point in the direction of the theory put forward herein. For example: "The nature of the mass motions in super giants probably resembles prominence activity", and "the motions observed in stellar atmospheres may be related to the ejection phenomenon", but these are all phenomena which only receive logically inter-related explanations on the discharge theory.
The explanation afforded by the theory for these broadened spectrum lines led to a theory of ball lightning(34.2, 35) which explains all the available data, and is supported by copious photographs of the lightning, long-spark and other laboratory discharges.
Fig.6: Comparison of rotational velocities of stars of Types B and Be (Slettebak).
(3.7) Low Stellar Atmospheric Density Gradients and Apparent Loss of Matter
In first emphasizing the possible use of the Doppler shifts of the lines in their spectra as the basis of a cosmic gas-velocity thermometer in 1956, it was pointed out(2.17) that this characteristic of electrical discharges would account for another well known characteristic of many stars, namely that the density gradients in their atmospheres are remarkably small; more than ten times less than can be accounted for when all account has been taken of all known factors, including turbulence.
A corollary to this general explanation of the physical process by which these atmospheres are extended by the discharge-generated gas jets is that the theory also throws a new light on those phenomena which have usually been regarded as demonstrating the continuous loss of matter from these stars in the outer atmospheres of which electrical discharges are occurring. The matter is not necessarily being lost to the star, but is merely being pushed further out in its atmosphere, which last will ultimately bring it to rest.
(3.8) g Cassiopeiae
(3.8.1) Rotating Four-Armed Nebula
McLaughlin(39), one of the leading authorities on these emission-line stars, five years ago wrote that "the behaviour of g Cassiopeiae violated almost all my generalizations" so its vagaries would appear to form a good test for the application of the discharge theory. A full account of this application will be found in a paper published in the following year(2.31), so only a few of the main agreements between theory and observation need be outlined here, for far from proving a difficult subject, this star provided a number of confirmations of the theory's predictions in addition to that discussed in Section (3.6.1). The general conclusion reached was that g Cassiopeiae is the central star of a four-armed nebula which rotates in a period of 4 years, the arms being in the plane of sight so that there is a partial eclipse of the star by one of the arms of the nebula every year, each of which reduces the star's light, by about a fifth of a magnitude(2.31) (Fig.6). From a study of the past behaviour of this star D.R. Barber has found evidence that outbursts occur in each pair of arms successively in an overall period of 11 years, the intervals between successive outbursts being 7 and 4 years respectively.
(3.8.2) Variation of Emission Line Widths
During each outburst, such as that studied in detail by Baldwin(40) from April, 1935 to November, 1938, there are discharges in each of one pair of arms and hence two diametrically opposed jets of hot gas. When the jets are in the line of sight, as happens every two years, the emission lines are broadened, the velocities of the jets being ±150 km/sec around J.D. 2428200. When the arms involved are at right angles to the line of sight, around J.D. 2428800, the two bright emission components merge into one undisplaced. narrow line but the star's luminosity is high(2.31) (Fig.5). The components again separate to reach velocities of ±120 km/sec around J.D. 2429100.
(3.8.3) Relation Between Intensities and Velocities of the Emission Line Components
The velocity indicated by the Doppler shift of each of the two components, R (red) and V (violet), should be linked with their intensities. Increase in the line intensity denotes an increase in the current in the discharge and in its temperature, which will lead to an increase in the velocity of the jet. That this expected relationship is observed is indicated by the Table:
|Intensity of R and V Components
|Velocity of R and V Components VR and VV
|R > V
|VR > VV
|V > R
|VV > VR
|R > V
|VR > VV
(3.8.4) Interchange of Line Components
The 1955 outburst was also studied by W.J.S.Lockyer(41) at the Norman Lockyer Observatory and it is interesting that one of his conclusions, which was subject to criticism, was that after the emission components had come together when the lines were at right angles to the line of sight, they had interchanged when they again separated, so that what had been the R component was now the V component, and vice versa. It will be seen that Lockyer's conclusion was correct.
(3.9.1) Line Broadening in Initial Stages
In presenting for publication the results described in Section (2.3.1) the writer emphasized(2.13) that the spectra of novae should show even greater line broadening than is observed in the spectra of solar flares, since the nova outburst is much greater than any solar flare. In accordance with this expectation it was recalled that the lines in the spectra of supernovae are so widened as almost to give the impression of a continuum.
It came as no surprise, therefore, to find that Stratton's measurements of the widths of the Balmer lines in the early spectra of Nova Herculis are proportional to the square of the wavelength, and not simply to the wavelength(2.36), i.e., their broadening is the result of Zeeman effect and not, as was usually supposed, of Doppler effect. When the note based on this observation was sent to "Observatory" one of the Editors wrote that it was interesting that the writer's theory led to the expectation of a broadening dl [proportional to] l2*, since this law had earlier been observed by Wright(42) in the early stages of N. Geminorum,1912. Though at that time Stratton was strongly opposed to the existence of the l2-law, in later years he actually advised the writer to "plug at your l2 problem", as no one had any alternative explanation to account for it.
(3.9.2) Current Wave-Shape
All atmospheric discharges will have similar current wave-shapes. A relatively rapid rise to peak current will be followed by a somewhat slower decline and a long tail, the form we are familiar with in the lightning discharge. This is for example the wave-form shown by the broadening of the Ha line in solar flares. It is also a well-known feature of the spectra during the initial stages of novae and efforts have been made by many investigators to explain it in terms of a rapid acceleration of the luminous gas followed by an almost equally rapid deceleration. As this part of the wave-form coincides with the period during which the line-broadening, dlal2, is that of the Zeeman effect, it is at once explained by the rise and fall of the current in the discharge.
(3.9.3) Pinching of the Discharge
Attention was recently directed to another interesting link-up between discharge phenomena on the terrestrial, stellar and galactic scales(2.37, 2.38), and this proved yet another example of the elucidation of terrestrial discharge phenomena from a study of their cosmic counterparts. Many novae, such as N. Aurigae, 1891 and N. Herculis, 1934 (Fig.7) have been practically extinguished soon after maximum magnitude or peak current. The light of both stars fell abruptly by nine magnitudes. This can only be accounted for by pinching out of the discharge. Photographic evidence of the occurrence of this phenomenon has been obtained in a lightning flash, and in a galactic discharge(2.30) (Fig.3).
This suggestion led to a search of the literature for oscillographic evidence of pinched lightning discharges which brought to light the oscillogram shown in Fig.8 (Ref. 2.39), in which the lightning current wave is shown on two different time scales. The cause of the pinching is discussed in a recent note(2.37) It results from the increase in pressure due to both the temperature rise and the aggregative force of the discharge. In the lightning discharge the combined effect will be to raise the pressure to well over 100 atmospheres. It has been found that the voltage gradient required to maintain a discharge increases as the square root of the pressure(2.40, 2.41) so that while the initial field may be adequate to maintain the leader stroke or the initial stages of the nova or galactic discharges, it may be inadequate to maintain the discharge in these later high pressure stages.
Fig.7: Light curve of Nova Herculis 1934.
Fig.8: Oscillogram of high lightning current wave (Berger and Vogelsanger). Curve shown on two time scales
(3.9.4) Two Discharges
In R Monocerotis (Hubble's Variable Nebula, Fig. 2a) and in others of the cometary nebulae the originating nova outburst had taken the form of a single discharge. Most planetary nebulae, however, as we have seen in Section (3.5.1) are two-armed as are their counterparts on a galactic scale. In an E.R.A. report in 1958 (Ref. 2.42) the writer explained how this will come about on the electrical discharge theory. Though there will in all probability be one initial and main discharge, it is quite probable that conditions will approach breakdown elsewhere in the stellar or galactic atmosphere, and that the ultra-violet and thermal radiation from the initial discharge will trigger off other discharges in both hemispheres. Those in the same hemisphere as the major discharge will be electromagnetically attracted to it; those in the opposite hemisphere will be repelled to the diametrically opposite point in the atmosphere.
In confirmation of this expectation Pearson(43) found that the observations on N. Aquilae, 1918 could best be explained on the hypothesis of two diametrically opposed gas jets and in the "Encyclopaedia of Astronomy" of Rudaux and de Vancouleurs it is stated(19.2) as a generalization that "In fact, it is known that the gases are not expelled uniformly by a nova, but primarily in two favoured and diametrically opposed directions". This is also confirmed by the detailed descriptions of individual novae given in Cecilia Payne-Gaposchkin's "Galactic Novae".
The main difficulties in the way of recognizing planetary nebulae as later stages of novae have been: (1) the much lower gas velocities in the nebulae, and (2) the relatively small number of planetary nebulae. However, the difference in the velocities is simply explained by the reduction in the discharge temperature toward the end of the discharge. The discharge-generated gas jet has become a pale shadow of that generated in the high current phase, when to judge from its velocity, the discharge temperature must on occasions have reached millions or hundreds of millions of degrees. Secondly, the final stage of a planetary nebula will be invisible, when the discharge ceases, so we do not expect to find as many planetary nebulae as there have been novae in our galaxy.
(4.1.1) Hubble's Scheme
The theory seemed to offer a satisfactory background on which Hubble's scheme of galactic evolution, Fig.9 (Ref. 44.1) could be explained. The atmospheric electric field is built up during stages E0 to E7; breakdown occurs at around Type S0 when the discharge channels form the spiral or barred spiral arms. We have seen that in a large proportion of the analogous stellar outbursts the discharges remain straight. In galaxies this is less common, though barred spirals, those in which the two discharges have remained straight for an appreciable proportion of their life, are about as common as spirals. If one discharge is deflected from the radial direction by a random aggregation of space charge to one side or the other ahead of it, then the change in the electrostatic field at the advancing tip of the other discharge will be such as to make it turn off in the other direction and so spiralling of both discharges is initiated.
It was disconcerting when for about a decade attempts were made to reverse Hubble's scheme of galactic evolution for reasons which the writer considered to be quite erroneous. However, it was reassuring to see that a movement back to Hubble's original scheme was initiated by Massevich(45) and others at a recent I.A.U. Symposium.
Fig.9: Hubble's sequence of nebular types.
(4.1.2) ) Type S0
Hubble wrote(44.2) that there is evidence for some "cataclysmic action" at S0. The occurrence of two discharges, the temperature of which reaches as we shall see four or five hundred million degrees, and which go on extending at 2,500 miles per second for ten or a hundred million years, certainly fulfils this requirement.
The discharges occur well inside the galaxy's dusty atmosphere, so that they have proved difficult to photograph. They only become generally visible as a result of the aggregation on to them of a large proportion of the galaxy's atmospheric matter after they have been long completed, which accounts for Hubble's repeated comment that the arms are never seen in process of formation but become visible as a whole at Type Sa or SBa (Ref. 44.2). However, this is no longer true. In view of the special interest of the apparently normal globular nebula NGC 4486 which is also the radio source Virgo A, efforts were made by careful filtering of its light to see what is going on, and the nearer of the two discharge channels was photographed. (See for example Ref. 2.30, Figs.2 and 3).
Fig.10: Diagram illustrating formation of Hubble's nebular type Sba.
(4.1.3) Type SBa
Hubble's galactic type SBa will be seen to be quite different from types SBb and SBc, but to resemble rather more those planetary nebulae in Fig. 4 in which the discharges have also remained straight. This form will be understood from Fig.10.
As the discharge approaches the perimeter of the star's or galaxy's spherical or rather oblate spheroidal or "elliptical" atmosphere the electric field ahead will be reduced but the lateral field will be high and the encircling magnetic field will be reduced, so that lateral breakdown, which this magnetic field had tended to prevent, will become possible. This will be seen to lead to a form having some similarity to the Greek letter q on both galactic and stellar nebular scales and the general breakdown throughout the peripheral regions of the whole atmosphere has given rise to the idea that the whole nebula has a spherical form. This impression is increased in a stellar (planetary) nebula if the discharges are viewed end on as in Fig. 11, when the effect is as though the nebula were a uniform ball of gas illuminated by the light of the central star.
Fig.11: Photograph of planetary nebula with arms in the line of sight.Note: The images of the nebulae from this original article was too poor to reproduce, and has been replaced by alternative images from another sources. The captions remain unchanged
(4.1.4) Radio Galaxies
When radio galaxies were discovered five years after the initiation of the discharge theory, it was immediately realized that they must be those galaxies in which the discharges are in progress, so that they must be either of type S0 or of late E types around E7. This conclusion has been confirmed by the statistical survey of the types of known radio galaxies discussed by Biermann and by Minkowski at a recent International Astronomical Union (I.A.U.) Symposium(46) from which 16 out of 24 radio galaxies are seen to be of type S0 or neighbouring types. It is also supported by the more recent investigations of Matthews, Morgan and Schmidt ("Quasi-Stellar Sources and Gravitational Collapse" Ed. Ivor Robinson, Alfred Schild and E.L. Schucking, University of Chicago Press, 1965, p.105).
(4.2) Atmospheric Electric Field Building
(4.2.1) Existence of Grains
In all the discussions of the building up of these cosmic atmospheric electric fields and in extenuation of their inadequacy at the present time, it should be kept in mind that an agreed answer to the question of the physical processes involved in the charge separation and field-building in thunderclouds a few miles above our heads is still lacking despite over 200 years of researches with kites, balloons, manned and unmanned, aeroplanes and radar. It is hardly to be expected that one can give a satisfactory explanation of the build-up of cosmic atmospheric electric fields at the present juncture.
We have seen, however, that the main requirement is dust or grains which can become charged and so act as electrodes applying the field to the gas. So it was significant that van de Hulst wrote "Where gas collects grains collect"(47). He went on to describe four main roles played by the grains in gaseous atmospheres, but in the writer's opinion omitted that role which far transcends all others in importance, they become charged on impact and can be separated in a gravitational field in which there exists another radial force such as radiation pressure or convection currents and so set up an electrostatic field. It is no general argument against the theory to say that space is infinitely conducting. It is also infinitely insulating, and if there are any free electrons they will be swept onto one of the grain electrodes. When all free electrons are swept out in this way there can be no further increase in the current until ionization by collision occurs and discharges are thus initiated.
The main question is, therefore, do the grains required by the theory exist in galactic atmospheres despite their extremely low densities -- one atom cm-3, as against the 109 molecules cm-3 in the earth's atmosphere? The ratio of these grains to gas atoms in a galactic atmosphere is actually a million times greater than the ratio of even the Aitken nuclei to molecules in our own atmosphere(48). So the electrodes are there to apply the field to the gas and the general coherent account of galactic phenomena, as well as stellar, to which the theory leads, offers fairly conclusive proof that the fields are built up to break down in electrical discharges. It is also known that the grains are needle-shaped and are oriented in one direction in one region of our galaxy as they would be were they dipoles oriented in an electric field, since they polarize star light coming from that region of the galaxy.
(4.2.2) Galactic Size -- Type Relation
There is, however, one check on the theory. It leads to the conclusion that small galaxies will evolve faster than will large galaxies, so that we should expect that small galaxies will pass relatively quickly along the evolutionary path from left to right (Fig.9) leaving the large galaxies behind among the E types. This expectation has been found to be fulfilled in recent years(49).
(4.2.3) Argument Against Continuous Creation
The explanation which the theory affords for the observed size-type relationship will be seen to afford a categorical argument against the theory of continuous creation(2.43). If past time had been unlimited then there would have been time to build up the electric field to break down even in the largest galaxies. The whole theory is based on a process of universal evolution, each step from universe to galaxies, galaxies to stars, stars to planets, being accelerated and governed by the electromagnetic aggregative force in the universal, galactic and stellar discharges. It is of interest, therefore, that the contradictory theory of continuous creation has had to be abandoned for other reasons as well.
(4.3) Discharge Channels
(4.3.1) Two-Armed Spirals
The explanation of the two-armed structure(2.42), first put forward to account for this characteristic of galaxies, has already been discussed in Section (3.5.1) since it equally well explains this structure on a stellar atmospheric scale.
Hubble's type SBa has already been discussed in Section (4.1.3). They are galaxies in which the discharges remained straight throughout the whole process of electrical breakdown(2.44). Types SBb and SBc are those in which the discharges departed from the radial direction sooner or later(2.45).
(4.3.3) Gas Jets in NGC 1097
One prediction immediately followed from the conclusion of the previous section(2.32) If the slit of a spectrograph is set along the bar of a barred-spiral galaxy, then if there are bright emission lines in its spectrum they ought to be displaced in opposite directions along the two halves of the bar owing to the outward flow of the discharge-generated jets of gas along each half of the bar. This has actually been observed in the barred spiral galaxy NGC 1097 by Burbidge and Burbidge (Fig.12) (Ref. 2.32, Fig. 7). Along each half of the bar the velocity is constant to within ±25 km/sec, but the two jet velocities differ by about 415 km/sec, a value which will depend on the inclination of the bar to the line of sight.
Fig.12: Two lines in the spectrum of the barred spiral galaxy NGC 1097 (Burbidge and Burbidge).
(4.3.4) Irregular Galaxies
At first it was not easy to understand why these galactic discharges remain in one plane till it was realized that just as there is a rotational diametral plane, so also will there be a diametral electrical plane in which the radial electric field will be a maximum(2.42). The neutralizing electrical discharges will be initiated in and confined to this plane, while lateral corona discharges will gradually neutralize the rest of the field.
However, if there is insufficient remnant angular momentum in the gas to cause it to show an oblate spheroidal or "elliptical" form, then there will be no preferred electrical plane either; the discharges will be governed by random accumulations of space charge and will be quite irregular. Apart from their shape they should otherwise be similar to those in spiral galaxies.
There are thus formed the 1% or so of irregular galaxies, which the theory prepares us to expect, which are similar as regards contents of arms and background to the more normal spirals and barred spirals. Their existence would seem entirely to preclude rotation as the root cause of the arm formation, though it is the basis of all other theories.
(4.3.5) Discharge Temperature
The temperature of these discharges cannot go on increasing indefinitely with their size and current(2.30). Ultimately those temperatures will be reached at which thermonuclear processes will be induced, when the rise in gas pressure due to the release of this energy in it will balance the electromagnetic aggregative force. This will presumably occur somewhere between 108°K and 109°K, so that the maximum velocity of the jets of gas formed by the discharges should lie between the velocity of sound in ionized atomic hydrogen at these two temperatures, i.e., between 1,750 and 5,400 km/sec.
Seyfert(50) has investigated the gas velocities in those galaxies which show an emission line spectrum denoting the occurrence of discharges(2.7) and has found velocities up to 4,500 km/sec in accord with these theoretical expectations. On the contrary the reason for the existence of this upper limit to cosmic gas velocities was otherwise merely the subject of a question at a recent I.A.U. Symposium(51) ; was it hydrodynamic? No, it is thermonuclear and clearly.
(4.3.6) Discharge Duration
The considerations in the preceding section led to a new theory of discharge propagation under these conditions(2.30), since the velocity of these jets of hot gas in the advancing leader stroke, 4.5 x 108 cm/sec, exceeds by a factor of more than ten that of voltage breakdown, about 3 x 107 cm/sec. When the discharge temperature reaches about 10,000,000°K the jet velocity will exceed this latter value and the jet will take over the propagation mechanism.
From this velocity of propagation and the total length of the spiral arms we can calculate the duration of the discharge -- and hence of the radio-galaxy-phase of a galaxy's life -- as about, 107 or 108 years, a value which agrees with that derived from statistical surveys of galaxies.
(4.3.7) Magnetic Fields in the Arms
The discharge theory of necessity introduced circular magnetic fields in and around the spiral arms from its initiation in 1941. Twelve years later Chandrasekhar and Fermi(52) introduced magnetic fields along the arms in an attempt to explain the energy of cosmic rays but as Cowling(53) emphasized in his "Magneto-hydrodynamics", this does not get us very far without an origin for these magnetic fields. After two papers based on Chandrasekhar and Fermi's fields Hoyle and Ireland(54) in a third in 1961 changed the fields to circles to agree with those the discharge theory introduced twenty years earlier and showed that such fields will equally well serve the theorist's purpose.
These suggestions concerning the magnetic fields, together with the changes suggested in the process of galactic evolution, Section (4.1.1) amply demonstrate the complete lack of knowledge of the physical processes involved and the need for such a theory as that discussed herein.
(4.3.8) Interacting Galaxies:
When a speaker who had suggested at an I.A.U. Symposium that the tilts in the galactic plane obviously caused in some way by the Magellanic Clouds might be a tidal effect, he was asked if they could be explained quantitatively in this way. His reply was illuminating(55): "... the distortion observed ... is too large by about two orders of magnitude to be explained by gravitational effects. But really I am not surprised. We see such enormous distortions in many galaxies and bridges between galaxies, and also more fancy things which cannot possibly be explained by gravitation, neither in order of magnitude nor even qualitatively in shape".
Figure 13 shows one of the "fancy things" which the speaker probably had in mind, for the effect of the upper on the lower galaxy is obviously not such as can be explained by gravitational effects. However, at least "qualitatively in shape", it is certainly in accord with what is to be expected on the discharge theory. For, if large currents flow in the spiral arms or discharge channels of the large upper galaxy, while the discharges were also in progress in the smaller lower galaxy, then the currents in the latter would be constrained to flow along the magnetic field due to the former and account for the right angled turns taken by the discharges in the lower galaxy.
(4.4) Stellar Populations
In the original short account of the work(2.7) it was suggested that the stars in a galaxy should be separable into two populations though so far as the writer was then aware, no one had previously suggested this. However, though this prediction was right, the reasoning was wrong. The correct explanation for the observed separation would, however, appear to have been given by thc discharge theory(2.42). The older stars of Population II are those formed from the primordial gas of the original galactic atmosphere contemporaneously with the building up of the electric field in the remainder of the gas. They are unaffected by the discharges when
they occur. The effect of the latter is to aggregate much of the remaining gas and dust on to the spiral arms. In these regions of considerably increased density a second population of younger stars is formed relatively quickly, Population I.
(4.4.2) Atomic Constitution
This suggestion is subject to numerical checks(2.47, 2.48). In the first place the gas which goes to the formation of Population I stars has been subjected to "thermonuclear" discharge temperatures of around 5 x 108°K for a period of the order of 107 years and its constitution must have been affected thereby. They should, therefore, contain a higher proportion of heavy atoms than do Population II stars despite the latter being by far the older. This difference is in fact observed: the heavy atom content of Population I stars is about 3%; that of Population II stars is only about 0.3%.
(4.4.3) Energy Radiated. by a Radio Galaxy
The effect of the discharges is two-fold: (1) during their propagation the galaxy acts as a radio galaxy; (2) they effect this change in the chemical constitution of the gas. In these two processes the amounts of energy involved must be the same. In other words during its phase as a radio galaxy an amount of energy is radiated equivalent to about 10-4 of its mass, since at out 3% of its matter loses about 1% of its mass. The mass of a galaxy is about 1043 to 1044 gm, so an amount of energy equivalent to about 1039 to 1040 gm, or about, 1060 to 1061 ergs, is radiated during the radio-galaxy phase. This estimate was higher than that available for comparison at the time, 1058 ergs (Ref. 56) derived from observations on the energy radiated by Virgo A, but has since been confirmed by Heeschen's fuller study of many radio galaxies, which yielded a value of 1060 ergs (Ref. 57).
(4.5) Optical and Radio Sources
The suggestion discussed in relation to solar flares in Section (2.3.5) is at once applicable to galactic discharges(2.24). It is to be expected that all these thermonuclear electrical discharges will be preceded in space by regions containing these non-thermal high speed electrons. Since the latter have velocities reaching at least 40 times the velocity of the H-ions and hence of the discharges themselves, we should expect these non-thermal extensions ahead of the visible discharges to reach dimensions of the order of 10 to 100 times those of the latter. Two quotations from a paper by Shain(58) to an I.A.U. Symposium on radio astronomy confirm these predictions of the discharge theory. About Virgo A he writes "It has been known for some time that at least the greater part of the radio emission from this source originates in a region about 5 minutes of arc in diameter and is associated with the main body of the nebula NGC 4486. More recently, Baldwin and Smith have given evidence for a faint extended source about, 50 minutes of arc in diameter surrounding the small bright source .... Observations now suggest that this conclusion needs modification. It appears that together with but not surrounding, the main "point" source there is a much fainter component, extending for between 0°.5 and 1°.0 on the preceding side ... The interest in this observation lies in the fact that the general direction of the extension is the same as that of the "jet" which optical observations have shown issuing from the nucleus of NGC 4486". This will be seen to coincide exactly with the theoretical expectations. Later in the same paper he writes, "when angular sizes of a number of radio sources are determined, any optical search for associated galaxies must extend to galaxies having diameters down to one-tenth or less of the radio diameters".
Just as in Section (2.3.5) theory and observations are thus in accord as to both the juxta-position of thermal and non-thermal sources or in this case optical and radio sources and as to their relative sizes.
(4.6.1) Their Absence a Theoretical Difficulty
In marked contrast to the impact which the discovery of quasars had on other theories, the absence of any reference to observations of such bodies had proved embarrassing to the discharge theory prior to their discovery(2.38), and the writer had had to conclude that their radiation must be largely in the unobservable regions of the spectrum, e.g., X-rays or g-rays.
The current wave-forms of all atmospheric electrical discharges must be similar. A rapid rise to peak current will be followed by a somewhat slower decline and a long tail. We see this wave-form delineated by the broadening and subsequent narrowing of the Ha line in the spectra of solar flares and novae, and in the rise and fall of brightness of the latter throughout the naked eye phase. Where were the "naked eye phases" of galactic discharges? There ought to have been radio galaxies radiating an inordinate amount of energy analogous to these naked eye phases of novae, but no one had reported them prior to the discovery of quasars.
(4.6.2) Optical Characteristics
In the original account of this work(2.7) the writer called attention to those galaxies which Hubble had found to be emitting an emission line spectrum and suggested that they probably fulfil the theory's requirement. These galaxies, which were studied in more detail by Seyfert(50), have turned out to be quasars, and this emission line spectrum, underlined in 1944, is still their only optical characteristic referred to in a recent review article on quasars(59).
It is of interest that the writer roughly estimated the duration of this phase of a galaxy's life in a paper presented to the second U.S.A.F. Conference on Atmospheric Electricity in 1958 (Ref. 2.6) still four years before their discovery. He pointed out that this phase of a lightning current wave lasts for m sec or tens of m sec of a solar flare for minutes and of a nova for days. In view of the similarity between the geometry of the nova and galactic outbursts we may assume that the peak current phase will last for similar proportions of the total life of the discharge in both cases. This leads to an estimate of the duration of the quasar phase of 104 to 105 years, which agrees well with Greenstein's(60) range of values of 103 to 107 years.
(4.6.4) Pinching of the Discharge
A quite unexpected characteristic of quasars is the rapidity with which their energy output varies. The writer has suggested that this may result from pinching of the discharge for which there is some evidence in the well-known photograph of Virgo A (Ref. 2.30, Fig.3), which shows that the discharge channel is pinchied into three "sausages". Analogous photographs have been obtained of a pinched lightning flash, and this prompted the writer to search for confirmation of such a process in published oscillograms of lightning current waves, with the satisfactory result(2.39) shown in Fig.8. It will be seen that the current rose rapidly to 180 kA, fell still more rapidly to a mere 10kA, i.e., the discharge was practically extinguished and rose again more slowly to about 50kA. An explanation for such occurrences has been given in terms of the writer's theory of the leader-return stroke relationship(2.49)
The corresponding phenomenon in the nova outburst was referred to in Section (3.9.5). (4.6.5) Association wish Dust
In view of the important role which atmospheric grains or dust have played in the theory since 1955, it is interesting that in the final section of a paper on "Theories of the Origin of Radio Sources" beginning "Let us collect together the Observational facts that a successful theory of radio sources has to account for", Burbidge and Burbidge(61) end by stressing "the role played by dust in radio sources". "The presence of dust", they go on, "in an otherwise normal-looking galaxy has come to be accepted as a good indication that a tentative identification with a radio source was in fact correct (M84 (Wade 1960) is an example). The broad dust lane in NGC 5128 has already been referred to. The dust pattern in the radio source NGC 1316 (Formax A) is another good example ... Among the Seyfert galaxies, a curious dust bar near the centre can be seen in NGC 3227. The very chaotic and remarkable dust structure in M82 is one of its most striking features".
This last confirmation may be one of the discharge theory's most striking features.
(4.7) Two Populations of Galaxies?
The theory suggested(2.7) that occurrences on a galactic scale have probably been preceded by a similar history of events on a universal scale and in particular that there may be two populations of galaxies corresponding to the two stellar populations in our galaxy(2.50). The one population of older galaxies would be forming contemporaneously with the building up of the universal electric field. The other of younger galaxies would be formed relatively quickly in the gas of increased density aggregated along the universal discharge arms.
This suggestion led to a search of the literature for indications that such collections of galaxies exist as the theory leads one to expect. The following facts are quoted from the "Larousse Encyclopaedia of Astronomy"(19.2). While the majority of the extra-galactic nebulae are apparently isolated in space "There is, however, no doubt that the Virgo cluster is a vast super-galactic organization of several hundred nebulae into a system whose shape is visibly flattened", while "the term 'clouds' is reserved for the much more extensive associations of galaxies whose existence was revealed by counts made at Harvard Observatory about 20 years ago. Among the most important of these clouds may be mentioned that which extends across Andromeda, Triangulum, Pegasus and Pisces and comprises several tens of thousands of faint nebulae". (The underlining is the writer's).
The theory thus predicts that the galaxies in these clouds will be found to be younger than the lone galaxies, but that the matter comprising them will be found to contain a higher proportion of heavy atoms than do the others, since it will have been subjected to thermonuclear temperatures during the universal discharges.
It is perhaps appropriate that this account should thus end with a prediction about which nothing can be said at present as regards its truth or otherwise.