the sun (figs. 53, 54, 55, and 56) or moon (figs. 57 and 58), but they are perfectly distinct from coronas, with which they should not be confounded. Halos are of comparatively rare occurrence; coronas, on the contrary, may be seen every time a light fleecy cloud comes between us and the sun or moon. The structure of halos, as seen from the figures, is often very complicated, circle cutting circle in the most remarkable manner, and with mathematical exactness, the diameters of the circles being generally very large; but the structure of the corona is simple, the circles concentric, the inner one small, varying from 2° to 4°, the diameter of the second circle being double that of the first, and of the third three times. In halos, the red prismatic colour is next the centre ; in coronas, the blue. Halos are formed from the refraction and reflection of the rays of light by the minute snow-crystals of the cirrus cloud; while coronas arise from the interference of the rays passing on each side of the globules of vapour.

627. Parhelia and Paraselence.-At the points of intersection of the circles of the halo, images of the sun or moon generally appear from the light concentrated at these points, the images of the sun being called parhelia or mock-suns, and those of the moon paraselenæ or mock-moons, which also exhibit the prismatic colours of the halo.

628. COLOURS OF CLOUDS. The gorgeous aërial landscapes of red and golden-coloured clouds which fire the western sky at sunset, "the day's dying glory" of the poet, all admire. They are observed to be the accompaniment of cumulus clouds (the cloud of the day during fine weather), while in the act of dissolving as they sink slowly down into the lower and warmer parts of the atmosphere, consequently they disappear from the sky shortly after sunset. Such sunsets are therefore universally regarded as prognostics of fine weather.

629. Frequently small thin clouds appear high up in the eastern sky some time before sunrise, or when

"The dappled dawn doth rise;"

and when the sun has risen they disappear. They are probably caused by the sun shining on and warming the upper layers of the atmosphere before it appears above the horizon; thus small ascending currents are formed, the vapour of which, as they ascend, is condensed in small clouds, or the cirro-cumulus. Their rounded definite forms show them to be produced in the same manner as the cumulus cloud-viz., by ascending currents forcing their way through colder strata. Their consistence is thin and vapoury, and their colour generally whitish or grey. They may thus be regarded as heralding the cumulus, and as sure prognostics of fine weather.

630. A green or yellowish-green tinted sky, on the other hand, is one of the surest prognostics of rain in summer, and snow in winter. An attentive consideration of the changing tints of the evening sky after stormy weather, supplies valuable help in forecasting the weather; for if the yellow tint becomes of a sickly green, more rain and stormy weather may be expected; but if it deepen into orange and red, the atmosphere is getting drier, and fine weather may be looked forward to.

631. Some years ago, Principal Forbes showed from experiments that high-pressure steam, while transparent, and in the act of expansion, readily absorbs the violet, blue, and part of the green rays, thus letting the yellow, orange, and red pass; and he suggested that coloration may be produced in a mixture of air and vapour, when the vapour is in the intermediate state referred to above. Dr E. Lommel has shown that successive layers of air with visible vapour diffused through them act, so to speak, like sieves, which continually separate the transmitted light more and more perfectly from its more refrangible rays. Hence, in passing through different thicknesses of vapour, the blue rays are first absorbed, then the yellow rays, and finally the red rays. It is in the lower layers of the atmosphere that dust, smoke, watery vesicles, and small rain-drops are chiefly suspended. When the sun is high in the heavens, the thickness of the vapour-screen between the sun and the eye is not sufficient

to produce any perceptible action on the rays of light, which consequently appear white; but as the sun descends to the horizon the thickness of the vapour is greatly increased, and at sunset it is calculated that the light of the sun has to pass through 200 miles of the air in illuminating a cloud a mile above the earth. Hence, as the rays fall more and more obliquely on the clouds, they appear successively yellow, orange, and finally red. The varied colours often seen at sunset are caused by the clouds appearing at different heights and in different parts of the sky, so that various thicknesses of vapour are interposed between them and the sun. At dawn the clouds first appear red; but as the sun rises higher, the yellow light ceases to be absorbed, and the clouds appear orange, yellow, and finally white. These successive stages of a perfect dawn are felicitously described in the 'Purgatorio :'

"The dawn was vanquishing the matin hour,
Which fled before it, so that from afar

I recognised the trembling of the sea.
Already had the sun the horizon reached,
So that the white and the vermilion cheeks
Of beautiful Aurora, where I was,

By too great age were changing into orange."

-Longfellow's translation.

Milton has accurately described the last stage of dawn in 'L'Allegro :'

"The great sun begins his state,

Robed in flames, and amber light

The clouds in thousand liveries dight."

632. It is evident that a high red dawn may be regarded as prognostic of settled weather, because the redness seen in clouds at a great height while the sun is yet below the horizon may be occasioned by the great thickness of the vapour-screen through which the illuminating rays must pass before reaching the clouds, and not to any excess of vapour in the air itself. But if the clouds be red and lowering in the morning, it may be accepted as a sign of rain, since the thickness traversed by the illuminating rays being now much

less, the red colour must arise from an unusual amount of vapour in the vesicular state, and in the state intermediate between the vesicular and the gaseous, when, according to Forbes, the blue rays are absorbed, and the yellow and red pass. It is to the latter of these kinds of red sunrise that the prognostic refers :

"The evening grey and the morning red,

Put on your hat, or you'll wet your head."

633. POLARISATION OF THE ATMOSPHERE.-The ordinary rays of light can, in certain circumstances, acquire new and peculiar properties, so that they cannot be reflected or refracted in the same way as before; they are then said to be polarised. When the light of any self-luminous body, such as the sun, is transmitted through certain crystallised substances, or reflected from, or refracted by, bodies not metallic, it is divided into two portions of polarised light, one of which is polarised in a plane at right angles to that in which the other is polarised. In doubly refracting crystals the two portions are polarised in opposite planes, and when common light is reflected from any body not metallic, whether solid, liquid, or gaseous, a portion of the incident light enters the body; and of the portions thus reflected and refracted, precisely the same quantity is polarised—the light polarised by refraction being polarised in a plane at right angles to that which is polarised by reflection. The polarising angle of a substance is the angle which the incident ray must make with the normal to a plane polished surface of this substance, so that the polarisation may be complete. The polarising angle for diamond is 68°; for glass, 57° 35'; for quartz, 57° 32'; for obsidian, 56° 30'; and for water, 52° 45'. The atmosphere acts upon light like other bodies, and consequently produces that physical change which constitutes polarisation. The polarising angle for air is

45° 0' 32".

634. The distance of the place of maximum polarisation from the sun undergoes great variations, chiefly from 88° to 92°. The general mean of an extensive series of observa

X

tions made by Dr Rubenson, Upsal, is 90° 2′, the half of which is so near the polarising angle of air as to leave no doubt that the light of the sky, as first stated by Brewster, is polarised by reflection from the particles of the air, and not from vesicles of water with parallel sides, as supposed by Clausius; nor, as conjectured by others, from minute drops of water, nor from molecules of aqueous vapour in an intermediate state between that of gas and that of vesicles.

635. Arago made the first great discovery respecting the polarisation of the atmosphere-viz., that there exists in the atmosphere a point or spot in which there is no polarisation. This neutral point is some distance above the point in the sky opposite to the sun, and which is called the antisolar point. The name of Arago's neutral point has been given to this spot of no polarisation, which is seen to best advantage after sunset. When the sun is in the horizon, Arago's point is about 18° 30' above the antisolar point in the opposite part of the horizon; but when the sun is 11° or 12° above the horizon, and the antisolar consequently as much below it, the neutral point is in the horizon, or it is 11° or 12° above the antisolar point. In abnormal circumstances it is sometimes only 7°, 8°, 9°, or 10°. As the sun descends to the horizon and the antisolar point rises, the distance of the neutral point from the latter gradually increases, and when the sun reaches the horizon the neutral point is, as stated above, 18° 30' above it. After the sun has set, the neutral point rises faster than the sun descends, and its maximum distance, when the twilight is very faint, is about 25°.

636. The second important discovery was made by M. Babinet in 1740. He discovered that there is a neutral point as far above the sun as Arago's neutral point is above the antisolar point. This spot is called Babinet's neutral point. It is best seen immediately after sunset, but is much fainter than Arago's point, being to some extent obscured by the yellow light of the setting sun.

637. Immediately after these discoveries, Sir David Brew