rays meet in one point. It more frequently happens, that they issue from a small mass which is of various shades of color, but generally grey or whitish; and this I have named the focus of radiation. In this mass

the rays melt away, or are lost. The convergence of rays to a single focus is sufficient to constitute a radiation : but as it is the nature of rays to stretch towards two opposite foci, it is probable, that the convergence towards the second focus

W

N

S

E

is prevented by circumstances, which do not render the phenomenon at all extraordinary. The radiation, consequently, will be only partial in one case, and entire in the other. If the rays be more apparent on one side than on the other, the radiation is unequal—and if the rays appear on one side only, it is then incomplete, or unilateral It is irregular when the rays are sinuous, or ill-disposed; direct

when the middle ray passes towards the zenith, while the lateral rays on each side have the same inclination in respect of the horizon.

It is inclined or oblique, when the rays on one side are nearer to the horizon than the cor

SSW

NNE

N

responding rays on the other side. I may remark, that sometimes, according to the degree of the obliquity, the two points of convergence are not directly opposite. Rays may be spread at unequal distances -and the separation, great or little, of the extreme rays may be denominated the amplitude of the radiation. The axis is a line which may be supposed to be drawn from one point of convergence to the other.

W

S

E

There are radiations, the rays of which are equal, or nearly alike. There are others, the rays of which are unequal, either in their dimensions, or in their consistency and color. For instance; on one side, they may be united, fine, light, or white; on the other side, thick, black, large, dappled and sometimes, even the lower are in well defined streaks.

We have hitherto spoken of radiation, as if the two points of convergence were exactly in the horizon. But it is not always thus; for both are often above the horizon, and sometimes they are below, though it more frequently occurs, that radiation is disposed in such a manner that one of them is above, and the other below.

I have often observed, that when the axis of the radiation is drawn from one equinoctial point to another, the two points of convergence are in the horizon, or equally elevated above the horizon; and that when the

axis is in the direction of north to south, or reciprocally, the focus or point of convergence on the north, is raised above the horizon, whilst that on the south is below. The disposition is rarely otherwise, and seems to favour the opinion of the ancients, that the winds blew downwards from the north, and upwards from the south. It is, however, unknown to me, whether that opinion was founded upon an exact observation of facts, or whether it was only a hypothetical consideration originating from the elevation of the pole in our northern climates. It is, at all events, certain, that, in general, the point of radiation is so much raised above the horizon as to correspond directly with due north; and if it be not placed in such a situation (which frequently happens) it is progressively lower in proportion as the focus approaches the east or the west.

Radiation is moreover a phenomenon so common, that few days pass without our having an opportunity of observing it. Radiations may be seen to succeed each other without interruption for some time. Nevertheless, there is no mention made of them in treatises of Meteorology. Mons. de Lamarck alone has, I think, in his Calender designated them by the name of Barred Clouds. But this illustrious author has paid but little attention to the characteristics of the phenoAs for me, I attach so much importance to it, that I regard it as the key to almost all the appearances of the celestial vault. *

mena.

That this phenomenon did not attract earlier atten

See a paper on this subject, with figures, by J. Rennie, Mag. Nat. Hist. vol. i.

tion, is much more astonishing, as it presents itself by night as well as by the day, under various and very beautiful forms. I could here produce from the writings of a great number of travellers ample testimony on this subject, and also in respect of the important appearance which, according to my conjectures on the Aurora Borealis, radiation makes in that brilliant meteor. Aurora," says Mons. de Chateaubriand, in his Itinerary from Paris to Jerusalem-" rose on our right above the lands of the Continent, on our left stretched the sea of Marmora ; the prow of our vessel was turned towards an island; the sky in the east was of a vivid red, which paled in proportion as the light increased. The scene

66

changed-numerous red and green rays, issuing from a common centre, ascended from the east to the zenith; these colors vanished, re-appeared, vanished again, until the sun rising above the horizon, confounded all the colourings of heaven in one universal whiteness, lightly tinged with gold."

I will now give an example of the formation, and of the different periods of radiation.—I will select one of the most simple cases, that can be met with. On the 19th of June, 1809, at three o'clock in the morning, the sky was very beautiful. But soon afterwards, a light cloud, streaked and not much elevated above the horizon, was formed to the W.N.W. This cloud became larger, and assumed the shape of a very well delineated arc, which stretched from the west to the north-east. It was of an iron-grey colour. The sun rose in great splendour beneath it. In the first, N. W.S. E. represent the entire circle of the horizon, and the arc is projected on the

C

plane of that circle.

The second represents

the appearance of the northern part of the sky; W.N.E. is a part of the horizon. The sun gradually rose to the height of the arc, by which it was eclipsed for some moments. The western part of the sky became insensibly dappled over, and the arc changed to a more

W

W

S

N

E

deep and sombre hue. The wind blew gently from the W.N.W. The dappled clouds were more dispersed, and progressively carried towards the region of the east, were formed into another mass. But, at the same time, the outline of another arc, similar to the former, was distinctly perceptible towards the south. The two small masses of dappled clouds uniting together occupied a space between the two arcs. It was now half-past five o'clock. The mass in the west had totally disappeared, while the east was become dappled all over. The arc, which extended to the south, was more developed. The wind from the W.N. W. blew more