Sir John Franklin mentions, in his ‘Journey to the Shores of the Polar Sea’:—“Nor could we distinguish its (the Aurora’s) rustling noise, of which, however, such strong testimony has been given to us that no doubt can remain of the fact.”
In detail, he mentions he never heard any sound that could be unequivocally considered as originating in the Aurora, although he had had an opportunity of observing that phenomenon for upwards of 200 nights (the Aurora was registered at Bear Lake 343 times without any sound being heard to attend its motions); but the uniform testimony of the natives and all the older residents in the country induced him to believe that its motions were sometimes audible. On the 11th March, at 10 P.M., a body of Aurora rose N.N.W.; and after a mass had passed E. by S., the remainder broke away in portions, which crossed about 40° of the sky with great rapidity. A hissing noise, like that of a bullet passing through the air, was heard, which seemed to proceed from the Aurora; but Mr. Wentzel assured the party the noise was occasioned by severe cold succeeding mild weather, and acting upon the surface of the snow previously melted in the sun’s rays. A similar noise was heard the next morning.
In Parry’s first voyage, Captain Sabine describes an Aurora seen at Melville Island, and adds that the Aurora had the appearance of being very near the party, but no sound could be heard.
In the article “Aurora Polaris,” Encyc. Brit, edition ix., the writer admits the evidence of scientific Arctic voyagers having listened in vain for such noises; but, referring to the statements of Greenlanders and others on the subject, concludes there is no à priori improbability of such sounds being occasionally heard, since a somewhat similar sound accompanies the brush-discharge of the electric machine.
Payer, of the Austrian Polar Expedition (1872-1874), states that the Aurora was never accompanied by noise, and discredits the alleged accounts of noises in the Shetlands and Siberia.
Lieut. Weyprecht, of the same expedition, says (antè, p. 14):—“Involuntarily we listen; such a spectacle must, we think, be accompanied with sound, but unbroken silence prevails, not the least sound strikes on the ear.”
Herr Carl Bock, who accompanied the Laplanders visiting this country (at the Westminster Aquarium) in 1877-78, and who witnessed many brilliant auroral displays in Lapland, assured me he could trace no noise, except on one occasion, when he heard a sort of rustling, which he attributed to the wind. The Laplanders themselves did not associate any special noise with the Aurora.
It has been recently stated, in an article on the Telephone in ‘Nature,’ that Professor Peirce “has observed the most curious sounds produced from a telephone in connexion with a telegraph wire during the Aurora Borealis;” but no further details are given. In experimenting with a silicic fluoride vacuum-tube between the poles of an electro-magnet, I found, on the magnet being excited, that the capillary stream of blue light was decreased in volume and brightness, and at the same time from within the tube a peculiar whistling or slightly metallic ringing sound was heard.
I certainly have never met with an instance of noise accompanying an Aurora and traced to it. On the whole the balance of evidence seems quite adverse to any proof of noises proper ordinarily accompanying an Aurora.
Colours of the Aurora.
Sir John Franklin considered the colours in the Polar Aurora did not depend on the presence of any luminary, but were generated by the motion of the beams, and then only when that motion was rapid and the light brilliant. The lower extremities, he says, quivered with a fiery red colour, and the upper with orange. He also saw violet in the former. Other observers have, in their various descriptions of Auroræ, mentioned the colours of the rays or beams as red, crimson, green, yellow, &c.; in fact, comprising the range of the spectrum. Violet seems less frequently mentioned. The red or crimson colour is frequently the first indication of the coming Aurora, and is usually seen on or near the horizon. The colours have frequently been observed to shift or change.
Prof. Piazzi Smyth, in a letter to ‘Nature,’ describing the Aurora of February 4th, 1872, as seen at Edinburgh, says that when the maximum development was reached all the heavens were more or less covered with pink ascending streamers, except towards the N., which was dark and grey—first by means of a long low arch of blackness, transparent to large stars, and then by the streamers which shot up from this arch, which were green and grey only for several degrees of their height, and only became pink as they neared the zenith. The red streamers varied from orange to rose-pink, red rose, and damask rose.
The Professor pointed out that the spectroscope knew no variety of reds giving one red line only, and attributed this to the mixing up of rays and streamers of blackness out of the long low arch. When the Aurora faded away a true starlight-night sky appeared; so that evidently the dark arch and streamers were as much part of the Aurora as the green and red lights.
Dr. Allnatt, at Frant, found in the case of the same Aurora the south-western part of the heavens tinged by a bright crimson band. A dark elliptical cloud extending from S. to S.E. was illuminated at its upper edge with a pale yellow light, and sent up volumes of carmine radii interspersed with green and the black alternating matter characteristic of elemental electricity. Almost due E., and of about 25 degrees elevation, was a bright insulated spot of vivid emerald-green, which appeared almost sufficiently intense to cast a faint shadow from intercepting objects. At 7 o’clock the Aurora had passed the zenith, and the sky presented a weird and wonderful appearance. A dark rugged cloud, some 8 degrees E. of the zenith, was surrounded by electric light of all hues—carmine, green, yellow, blood-red, white, and black; and the bright spot still existed in the south.
At Blackburn, in Lancashire, the rays were described as glowing in the N.E. from silvery white to deepest crimson; and at Cambridge the same Aurora was described as of a brilliant carmine tint. The Auroræ seen in Lapland by Herr Carl Bock, were, he informed me, almost invariably yellow; he saw only one red one.
The behaviour of a hydrogen Geissler vacuum-tube will be subsequently referred to in the Chapter on the comparison of some tubes with the Aurora spectrum, and is suggestive as to Aurora colours.
The capillary part of this tube, when lighted by a small coil, was found to vary in tint—silver-white, bright green, and crimson being each in succession the dominant colour, according to the working of the break of the coil. When a spectroscope was used, the red, blue, and violet lines of the gas were seen to change in intensity in accordance with the light colour seen in the tube.
A Geissler nitrogen vacuum-tube was also so arranged that the capillary part of it should be vertically between the conical extremities of the armatures of a large electro-magnet, the armatures just being clear of the outside of the tube. The tube was then lighted up by a small coil, and the magnet excited by four large double-plate bichromate cells.
The stream of light was steady and brilliant, and, except at the violet pole, of the rosy tint peculiar to a nitrogen vacuum-tube. On excitation of the electro-magnet, the discharge was seen to diminish in volume, with an apparent increase in impetuosity; and not only the capillary part, but in a less degree the bulbs also of the tube, changed from a rosy to a well-marked violet hue.
We several times connected and disconnected the magnet with its batteries, but always with the same result. Of the spectrum of the capillary part of this tube we took photographic plates with quartz prisms and lenses, taking care that all things should be as equal as possible, the apparatus undisturbed, and the time of exposure exactly the same. One plate was taken with the tube in its normal condition, the other while it was under the influence of the magnet. The spectra were identical, except that the plate of the tube influenced by the magnet was decidedly the brightest, and was found to penetrate more into the violet region (the Author’s ‘Photographed Spectra,’ p. 60, plate xxv.). These plates effectually corroborated the change of colour, as the violet ray would have more photographic effect than the rosy. The identity of the spectra of the capillary part proved that the change in colour could not have proceeded from an extension of the violet glow. (A similar experiment will be found also detailed in Part III. Chapter XII.)
Height of the Aurora.
Sir John Franklin (Narrative of a Journey on the Shores of the Polar Sea in the years 1819, ’20, ’21, ’22) says:—“My notes upon the appearance of the Aurora coincide with those of Dr. Richardson in proving that that phenomenon is frequently seated within the region of the clouds, and that it is dependent in some degree upon the cloudy state of the atmosphere.” And further:—“The observations of Dr. Richardson point particularly to the Aurora being formed at no great elevation, and that it is dependent upon certain other atmospheric phenomena, such as the formation of one or other of the various modifications of cirro-stratus.”
Sir John Franklin also refers to notes from the Journal of Lieut. Robert Hood, R.N., on an Aurora:—
The observations were made at Basquian House, and at the same time by Dr. Richardson at Cumberland House, quadrants and chronometers having been prepared for the purpose. On the 2nd April the altitude of a brilliant beam was 10° 0´ 0″ at 10h 1m 0s at Cumberland House. Fifty-five miles S.S.W. it was not visible. It was estimated that the beam was not more than 7 miles from the earth, and 27 from Cumberland House. On the 6th April the Aurora was for some hours in the zenith at that place, forming a confused mass of flashes and beams; and in lat. 53° 22´ 48″ N., long. 103° 7´ 17″, it appeared in the form of an arch, stationary, about 9° high, and bearing N. by E. It was therefore 7 miles from the earth.
On the 7th April the Aurora was again in the zenith before 10 P.M. at Cumberland House, and in lat. 53° 36´ 40″ N., long. 102° 31´ 41″. The altitude of the highest of two concentric arches at 9h P.M. was 9°, at 9h 30m it was 11° 30´, and at 10h 0m 0s P.M. 15° 0´ 0″, its centre always bearing N. by E. During this time it was between 6 and 7 miles from the earth. [The bearings are true, not magnetic.]
Sir J. Franklin says this was opposed to the general opinion of meteorologists of that period: he also noticed he had sometimes seen an attenuated Aurora flashing across the sky in a single second, with a quickness of motion inconsistent with the height of 60 or 70 miles, the least that had hitherto been ascribed to it.
The needle was most disturbed, February 13, 1821, P.M., at a time when the Aurora was distinctly seen passing between a stratum of cloud and the earth; and it was inferred from this and other appearances that the distance of the Aurora from the earth varied on different nights. Dr. Richardson concludes that his notes prove, independent of all theory, that the Aurora is occasionally seated in a region of the air below a species of cloud which is known to possess no altitude; and is inclined to infer that the Aurora Borealis is constantly accompanied by, or immediately precedes, the formation of one or other of the forms of cirro-stratus.
Captain Parry observed Auroræ near the earth’s surface; and records that he and two companions saw a bright ray of the Aurora shoot down from the general mass of light between him and the land, which was distant some 3000 yards. Sir W. R. Grove (‘Correlation of Physical Forces’) saw an Aurora at Chester, when the flashes appeared close, so that gleams of light continuous with the streamers were to be seen between him and the houses—“he seemed to be in the Aurora.” Mr. Ladd, of Beak Street, Regent Street, has related to me an appearance he was struck with, and examined carefully. Standing in the evening in Margate Harbour, he saw a white ray of the Aurora, which, apparently shooting downwards, was clearly placed between his eye and the opposite head of the pier, which projected into the sea. Mr. Ladd also informed me that Prof. Balfour assured him that such an appearance was not unusual. In the double-arc Aurora seen by me in the Isle of Skye, September 11, 1874 (described antè, p. 23), I had a strong impression that the bow was near the earth, and thought that the eastern end, and some fleecy clouds in which it was involved, were between myself and the peaks of the distant mountains.
In the article “Aurora Polaris,” Encyc. Brit., edition ix., Dalton is instanced as having calculated the height of an Aurora in the north of England at 100 miles; and Backhouse as having made many calculations, with the result of an average height of 50 to 100 miles. Prof. Newton, too, is quoted for the height of 28 Auroræ (calculated by one observation of altitude and amplitude of an arch) as ranging from 33 to 281 miles, with a mean of 130 miles. It is, however, pointed out that a height of 62 miles above the earth’s surface would imply a vacuum attainable with difficulty, even with the Sprengel pump. This difficulty is then met by a reference to the observed altitude of some meteors, and to a suggestion of Prof. Herschel’s that electric repulsion may carry air or other matter up to a great height. Dr. Lardner (‘Museum of Science and Art,’ vol. x. p. 192) speaks of the height of Auroræ as not certainly ascertained; but considers them atmospheric phenomena scarcely above the region of the clouds, and does not think it probable that their elevation in any case can exceed a few miles.
M’Clintock, after noticing that the beams of the Aurora were most frequently seen in the direction of open water, says that in some cases patches of light could be plainly seen a few feet above a small mass of vapour over an opening in the ice. Captain Ross, in his Antarctic voyage, saw the bright line of the Aurora forming a range of vertical beams along the top of an ice-cliff; and suggested this was produced by electrical action taking place between the vaporous mist thrown upwards by the waves against the berg, and the colder atmosphere with which the latter was surrounded.
Bergman, from a mean of 30 computations, makes the height of the phenomenon to be 72 Swedish (about 468 English) miles.
Father Boscovich calculated the height of an Aurora Borealis observed on the 16th December, 1727, to have been 825 miles.
Mairan supposed the far greater number of Auroræ to be at least 600 miles above the surface of the earth. Euler assigned them an elevation of several thousands of miles. Dr. Blagden, however, limited their height to about 100 miles, which he supposed to be the region of fireballs—remarking that instances were upon record in which northern lights had been seen to join and form luminous balls, darting about with great velocity, and even leaving a train behind them like common meteors (Phil. Trans. vol. lxxiv. p. 227).
Mr. Dalton, from an observation of the luminous arches on a base of 22 miles, found the altitude of the Aurora to be about 150 miles (Dalton’s ‘Meteorological Observations and Essays,’ 1793, pp. 54, 153).
Dr. Thompson, ‘Annals of Philosophy,’ vol. iv. p. 429 (1814), assumes that the height of the beams above the surface of the earth was much greater than that of most other meteorological appearances, and gives (p. 430) a table of Auroræ, mainly taken from Bergman, Opusc. v. p. 291, of 31 Auroræ observed in the years 1621 to 1793, with heights in English miles. The lowest is, 23rd February, 1784, London (Cavendish), 62 miles; the highest, 23rd October, 1751, Fournerius, 1006 miles! The average of the 31 estimated observations gives a height of about 500 miles. It is not stated how these observations were obtained, though methods are mentioned how they might be.
Prof. Heis, of Münster, exhibited at the recent Scientific Loan Collection at South Kensington (‘Official Catalogue,’ 3rd edit. p. 296, No. 1231) an instrument for the determination of the position of the point of convergence of the rays of the Aurora, and for determining the height of the Aurora. A ball resting in a pan was to be brought into position, so that several diverging pencils of Aurora, when properly viewed, were covered by the rod which passed through the centre of the ball. The point of the rod (which could be moved up and down in the ball), when the instrument was set to the astronomical meridian, showed the azimuth and altitude of the converging point of the pencils of light. This point of convergence does not coincide with the point to which the inclination-needle directs. From the deviation of the two points, the height of the Aurora could be calculated.
Professor H. A. Newton (Sil. Journ. of Science, 2nd ser. vol. xxix. p. 286) has proposed a method of calculating the height of Auroræ by one observation of altitude and amplitude of an arch. It assumes that the auroral arches are arcs of circles, of which the centre is the magnetic axis of the earth, or at least that they are nearly parallel to the earth’s surface, and probably also to the narrow belt or ring surrounding the magnetic and astronomical poles. Professor Newton finds that, d being the distance from the observer to the centre of curvature of the nearest part of this belt (for England, situated about 75° N. lat., 50° W. long.), h the apparent altitude of the arch, 2a its amplitude on the horizon, x its height, R the earth’s radius, and c the distance of the observer from the ends of the arch:—
| sin φ = sin d cos a cosec(d + h) | (1) |
| tan c = z sin h sin φ sec ²φ | (2) |
| x = R - (sec c - 1) | (3) |
This method with 28 Auroræ gave a height from 33 to 281 miles and a mean of 130 miles.
Galle has suggested (Pogg. Ann. cxlvi. p. 133) that the height of Auroræ might be calculated from the amount of divergence between the apparent altitude of the auroral corona and that indicated by the dipping-needle, a principle which has been adopted in Prof. Heis’s apparatus before described. The results do not differ materially from Professor Newton’s.
The conclusions to be arrived at from the foregoing instances and opinions are certainly very puzzling. The terrestrial character of some Auroræ seems well established. The height to which these phenomena may ascend is left almost a matter of conjecture, and further observations are very desirable.
Phosphorescence.
In the voyage of the ‘Hansa’ (‘Recent Polar Voyages,’ p. 420), on the 9th September, 1869, at 10 P.M., Aurora gleams appeared in the west, shooting towards the south. “Radiant sheaves and phosphorescent bands mounted towards the zenith,” but the phantasmagoria quickly vanished. M. Silbermann (‘Comptes Rendus,’ lxviii. p. 1120) mentions storm-clouds which threw out tufts of cirri from their tops, which extended over the sky, and resolved into, first, fine, and afterwards more abundant rain. (I saw a fine day example of this on the Lago di Guarda, ending in a copious discharge of rain attended with loud thunder and vivid lightning.) Usually the fibres were sinuous; but in much rarer cases they became perfectly rectilinear and surrounded the cloud like a glory, and occasionally shone with a sort of phosphorescence. On the night of 6th September, 1865, at 11 P.M., a stormy cloud was observed in the N.N.W., and lightning was seen in the dark cumulous mass. Around this mass extended glories of a phosphorescent whiteness, which melted away into the darkness of the starry sky. Round the cloud was a corona, and outside this two fainter coronæ. After the cloud had sunk below the horizon the glories were still visible.
Sabine mentions a cloud frequently enveloping Loch Scavaig, in Skye, as being at night perfectly self-luminous, and that he saw rays, similar to those of the Aurora, but produced in the cloud itself. Sabine also refers to luminous clouds mentioned in Gilbert’s Annals, and to observations by Beccaria, Deluc, the Abbé Rozier, Nicholson, and Colla; and to luminous mists as observed by Dr. Verdeil at Lausanne in 1753, and by Dr. Robinson in Ireland.
He also describes (Parry’s First Voyage) an Aurora seen at Melville Island, and says the light was estimated as equal to that of the moon when a week old. Besides the pale light, which resembled the combustion of phosphorus, a slight tinge of red was noticed when the Aurora was most vivid; but no other colours. This Aurora was repeatedly seen on the following day.
Mr Procter, in a letter to me, suspects that the Aurora is generally formed in a sort of “mist or imperfect vapour;” and this mist or imperfect vapour seems in many instances to form part of the Aurora, and to partake of its self-luminous character. M’Clintock does not imagine that the Aurora is ever visible in a perfectly clear atmosphere. He has often observed it just silvering or rendering luminous the upper edge of low fog or cloud-banks, and with a few vertical rays feebly vibrating.
An instance of apparent phosphorescence is supplied by the Aurora of the 4th February, 1874 (antè), when a bright cloud of light was seen which gave the impression of an “illuminated fog-cloud.” Captain S. P. Oliver saw at Buncrana, Co. Donegal, on February 4, 1874, what he describes as a meteor-cloud, viz. “a broad band of silvery white and luminous cloud.” This appearance, as described by another correspondent, was evidently an imperfectly formed (perhaps actually forming) Auroral arc. The great Auroral display of the 24th of October, 1870, as seen by me, included, according to my notes made at the time, “streamers of opaque white phosphorescent cloud, very different from the more common transparent Auroral diverging streams of light.”
Describing the Aurora of February 4, 1872, at Frant, Dr. Allnatt says:—“At a later hour of the night the canopy of cirro-stratus had separated, and was transformed into luminous masses of radiant cumulus. At 10.40 the Aurora reappeared in the N., and sent luminous radii of white phosphorescent light from the periphery of a segment of a perfectly circular arch”[7].
Again, February 4th, 1872, as described by me, the first signs of the Aurora were (in dull daylight) a lurid tinge upon the clouds, which suggested the reflection of a distant fire; while scattered among these, “torn and broken masses of white vapour having a phosphorescent appearance” reminded me of a similar observation in October 1870.
The day Auroræ, which are elsewhere described, and are not very uncommon, could, we may presume, hardly be seen without the presence of some phosphorescent glow. Professor Ångström, in his Aurora Memoir (discussed elsewhere), in discussing the yellow-green line, considers the only probable explanation to be that it owes its origin to fluorescence or phosphorescence. He says that some fluorescence is produced by the ultra-violet rays; and adds, “an electric discharge may easily be imagined, which, though in itself of feeble light, may be rich in ultra-violet light, and therefore in a condition to cause a sufficiently strong fluorescent light.” And he refers to the fact that oxygen and some of its compounds are phosphorescent.
In the examination of certain spectra connected with the Aurora, detailed in Part II., I have shown that the bright edge of one of the phosphoretted hydrogen bands is in close proximity to the yellow-green Auroral line. I have also referred to the peculiar brightening by reduction of temperature of one of the bands in the red end of the spectrum of phosphoretted hydrogen, so that from almost invisible it became bright, and to the peculiar brightening of a line in the yellow-green in certain “Aurora” and phosphorescent tubes. It has also been observed that the electric discharge has a phosphorescent or fluorescent after-glow (isolated, I believe, by Faraday). It seems difficult to avoid in some way connecting all these circumstances with the yellow-green line of the Aurora, if not also with the line in the red.
Mr. Sorby, in his experiments on the connexion between fluorescence and absorption (‘Monthly Microscopical Journal’), found in the spectrum of a solution in alcohol of a strongly fluorescent substance called bonelleine (the green colouring-matter found in the Aurelia Bonellia-viridis) two bright bands, the one red and the other green, with centres respectively at 6430 and 5880, and their limits towards the blue end at 6320 and 5820. On adding an acid the red band changed its place to 6140. The superficial membranous coloured layer of the fungi Russula nitida and vesca in alcohol gave an absorption band with centre at 5540, while the spectrum of fluorescence extended to 4400. A solution of Oscillatoriæ in water gave a spectrum of absorption with bands at 6200 and 5690; while the spectrum of fluorescence showed two bright bands having their centres at 6470 and 5800, and their limits towards the blue end at 6320 and 5710.
These instances of course cannot be connected with the Aurora except as showing the spectrum region and lines of fluorescence. The sea phosphorescence, according to Professor Piazzi Smyth, has a continuous spectrum extending from somewhat below E to near F (Plate V. fig. 3).
Ångström, on the occasion of the starry night when he found traces of the green line in all parts of the heavens, speaks of the sky as being “almost phosphorescent.”
The author of the Aurora article in the Encyc. Brit. suggests that the phosphorescent or fluorescent light attributed to the Aurora may be due to chemical action. He also questions Ångström’s assumption that water-vapour is absent in the higher atmosphere, and thinks that it and other bodies may, by electric repulsion, be carried above the level they would attain by gravity. He then continues that if discharges take place between the small sensible particles of water or ice in the form of cirri (as Silbermann has shown to be likely) surface decomposition would ensue, and it is highly probable the nascent gases would combine with emission of light. He adds “that it has been almost proved that in the case of hydrogen phosphide the very characteristic spectrum (light?) produced by its combustion is due neither to the elements nor to the products of combustion, but to some peculiar action at the instant of combination; and it is quite possible that under such circumstances as above described water might also give an entirely new spectrum.” Professor Herschel has referred to the phosphorescent light which remains glowing in Geissler tubes after the spark has passed, and to the fact that one of the globes of a “garland” tube which was heated did not shine after the spark had passed, apparently because of the action of heat on the ozone to which the phosphorescence might be due. (See experiments on Mr. Browning’s bulbed tube, Part III. Chap. XV.)
Aurora and Ozone.
Accounts are given by travellers in Norway of their being enveloped in the Aurora, and perceiving a strong smell of sulphur, which was attributed to the presence of ozone. M. Paul Rollier, the aëronaut, descended on a mountain in Norway 1300 metres high, and saw brilliant rays of the Aurora across a thin mist which glowed with a remarkable light. To his astonishment, an incomprehensible muttering caught his ear; when this ceased he perceived a very strong smell of sulphur, almost suffocating him (‘Arctic Manual,’ p. 726).
In the case of the Aurora, the question naturally arises whether the oxygen of the air may be changed into ozone, perhaps also whether the nitrogen may not be modified in some similar manner.
The absorption spectra of oxygen, and of the same gas in its form of ozone, may possibly differ; but this can hardly happen in the case of incandescent oxygen, for ozone is at once destroyed by heat at 300°, and slowly at 100°, and must be partially at least destroyed by the heat of the discharge. If any lines were due to ozone in such a spectrum, we should expect they would be weakened by heat and brightened by cold.
In the case of a continued discharge in a large exhausted bell-receiver, the presence of ozone in considerable quantities was manifested to us by its odour when the receiver was removed from the pump; but the spectrum of the stream of light did not appear to differ from that in Geissler tubes.
In a course of lectures at the Royal Institution in March 1878, on the Chemistry of the Organic World, Prof. Dewar appears to have demonstrated, by Prof. Andrews’ apparatus, that ozone is really condensed oxygen, and, further, that during this condensation heat is absorbed, which is evolved during the decomposition or re-expansion.
He also exhibited the oxidizing power of ozone in its action on mercury, and commented on its similar action upon organic matter in forming nitrates, and on its remarkable bleaching properties, but added there was as yet no proof of its combining with free nitrogen. That peroxide of hydrogen accompanies the formation of ozone by the slow combustion of phosphorus, and that this peroxide acts with ozone in decomposing organic bodies, though in an inexplicable manner, the Professor considered to be proved. He also referred to the silent discharge probably perpetually going on between the upper and lower strata of the atmosphere, and also between these and the earth, accounting, as the Professor considered, for some of the chemical actions whereby nitrogenous compounds are formed in the soil.
As far as I am aware, no information as to a possible spectrum of ozone, or a modification of the oxygen or other spectra by its presence, has, up to the present time, been obtained[8].
It has been suggested by Mr. Procter and myself that the electric discharge in an exhausted moist tube, if subjected to a considerable degree of cold, might produce a modification of the air-spectrum, perhaps even a spectrum analogous to that of the Aurora.
For some further notes on this subject see Appendix D (Aurora and Ozone).
Polarization of the Aurora Light.
In ‘Nature,’ vol. vii. p. 201, is contained an account of observations of the polarization of the zodiacal light and of the Aurora, by Mr. A. Cowper Ranyard, who, using both a double-image prism and a Savart on the great Aurora of February 4th, 1872, detected no trace of polarization. He also examined a smaller one of 10th November, 1871, with a like result.
Mr. Fleming (who refers to these observations) remarks that the only other account he had met with was contained in Prof. Stephen Alexander’s Report on his Expedition to Labrador, given in Appendix 21 of the U.S. Coast Survey Report for 1860, p. 30. Professor Alexander found strong polarization with a Savart’s polariscope, and thought that the dark parts of the Aurora gave the strongest polarization. This was in latitude about 60°, at the beginning of July, and near midnight. It is not stated whether there was twilight or air-polarization at the time, nor is the plane of polarization given.
The question naturally arises, especially as the darkest parts of the Aurora are usually situated low down near the horizon, whether the polarization in the latter case did not proceed from the atmosphere and not from the Aurora itself. Mr. Shroeder found no traces of polarization in the Aurora of February 4th, 1872. Further examinations of the Aurora with some delicate form of polariscope would seem very desirable.
The evidence of polarization in the case of the zodiacal light seems also almost entirely negative—Mr. Ranyard pointing out observations of his own, of Captain Tupman, and of Mr. Lockyer with this result. Mr. Burton, using a Savart set so as to give a black centre when the bands were parallel to the plane of polarization, believed he detected faint traces of polarization in the brightest parts of the zodiacal light (as seen in Sicily), the bands being black-centred when their direction coincided with the axis of the cone of light. Mr. Burton saw no trace of bands when examining the slight remaining twilight apart from the zodiacal light. Mr. Ranyard was not able to confirm Mr. Burton’s observations on the same evening and with the same instrument.
Number of Auroræ.
Sir John Franklin saw in the Arctic Regions, in the years 1819, 1820, 1821, 1822:—In the month of September two Auroræ, in October three, in November three, in December two, in January five, in February seven, in March sixteen, in April fifteen, and in May eleven.
Periodicity as to days seems to have no certain law; and though certain days in February and March are marked as those of fine returning displays, they must be looked on as accidental.
Two well-marked annual maxima seem to occur in March and October (the latter the greater), and two minima in June and January, the greater in June (Encyc. Brit.). The 4th of February, 1872, and same day 1874, are, however, curious instances of a recurring remarkable display.
A table by Kæmtz, showing the number of Auroras in each month of the year, with the maxima and minima as above stated, will be found on Plate V. fig. 5.
Dr. Hayes has observed that in the winter of 1860-61 (when the ten or eleven years’ inequality was at its maximum) only three Auroræ were seen and recorded, and they were feeble and short in duration.
Captain Maguire, at Point Barrow (1852-54), reports that the Aurora was seen six days out of seven, and on 1079 occasions, being nearly one third of the hourly observations. It was seldom seen between 9 A.M. and 5 P.M., not at all between 10 A.M. and 4 P.M. It increased regularly and rapidly from 5 P.M. until 1 A.M., and then diminished in the same way until 9 A.M.
The winters of 1877 and 1878 and the springs of 1878 and 1879 have been singularly deficient in Auroræ. I have seen none at Guildown.
Duration of Aurora.
In the article in the ‘Edinb. Encyc.’ before referred to some remarks are made on the duration of the Aurora. Sometimes it is formed and disappears in the course of a few minutes. At other times it lasts for hours or during the whole night, or even for two or three days together. Musschenbroek observed one in 1734 which he considered to have lasted ten days and nights successively, and another in 1735 which lasted from the 22nd to the 31st March.
With respect to Captain Maguire’s observations (antè) it may be remarked that Auroræ may doubtless frequently run on into and through the day without their being noticed (instances, however, are known of Auroræ seen in daylight); and hence it is difficult to judge of the limit of duration of a particular Aurora unless indications are sought for during the day (by the shapes of clouds, action of the magnet, &c.) as well as during the night. Probably Auroræ seen during successive nights may be parts of a continuous discharge.
The Travelling of Auroræ.
Donati undertook to study the Aurora with reference to the mode of its extension; and he arrived at the result that the Aurora of February 4, 1872, was not observed in different regions of the earth in the same physical moment; but everywhere at the same local hour, as in the case of celestial phenomena, which do not share in the earth’s rotation.