water-mark and the first cliffs inland, forms a flat space of greater or less breadth, along which the coast-road lies. This is most generally covered deep with coral sand, which in some places, as between Kahuka and Laie, forms elevated downs along the coast, which conceal the sea from the view of the traveller on the road. The specimens of this series are Nos. 31, 32, 52, 53. One of the best places for seeing a section of these cliffs is just at the north point of the island, above Kahuku; it there forms a precipice elevated at its summit about fifty feet above the level of the sea. The limestone forming these cliffs is of two kinds. One is loose and cellular, of a straw-yellow colour, is easily decomposed by atmospheric agency; when examined with a glass, it appears to be an aggregate of white botryoidal-shaped fragments, and is covered with an ochry looking layer of colouring matter. The other is harder and more compact, affords more resistance to the weather, and projects from the cliff beyond the former; it is of a yellowish-grey colour, with a small granular foliated, partly, and splintery fracture. It is singular that the latter, which appears the older rock of the two, in mineralogical characters, should be superposed to the former; it forms a tabular mass, running along the whole of the upper edge of the cliff, giving to it that prominence and steepness which renders it a conspicuous object for several miles on the approach from Laie, on the south-east. These coral cliffs, for such I would call them on account of the close similarity between the looser variety and the coral reefs, now forming at the water's edge, form extensive tabular cliffs for some distance round the north point of Oahu at a nearly uniform elevation.

The sand forming the extensive plains and downs at the base of these cliffs, is of a dazzling whiteness. After being kept in paper for some time, it has nearly a smoke-grey colour, and consists of variously coloured particles, varying in size from fine dust to that of a millet-seed, made up of fragments of shells, microscopic shells, comminuted limestone, and portions of a dark coloured rock, probably lava.

The soil of Oahu is of two kinds, either sand or a deep black earth, arising from the decomposition of lava. Both are so very porous that where there are not abundant facilities of irrigation, the barrenness is extreme. The taro fields are almost all in

hollows, with the deep lava soil at bottom, and a running stream in the middle. At first sight the lava soil appears to be a fertile black mould, but it is nothing more than an accumulation of round gravelly grains of lava rock.

Stratification. There were tendencies to this in some craters at the east end of the island, and at the plain of Eva, near Pearl River, at the salt lake; but these I had no opportunity of examining. The only distinct traces that I observed were on the north-east coast, near Waihea. I there met with specimens Nos. 13, 17, and 29. It consists of an alternation of beds of dark coloured compact basalt rock, from one to six feet thick, alternating with still thicker beds of grey colour, tuffaceous and amygdaloidal rock, different from any seen in other parts of the island. One of the beds (No. 17,) could not be distinguished from some kinds of greywacke slate in texture and colour of the basis; it contains disseminated numerous white points, which are probably crystals of calcareous spar. No. 14, is not unlike some kinds of wacke. The basis of Nos. 15, 16, 29, is nearly the same as 14, but contains larger amygdaloidal cavities. The whole of this series of rocks is directed N. 55° W. by compass, nearly parallel with the general direction of this side of the island, and the dip to the north-east, rising to the centre of the island. One of the most likely modes of accounting for the formation of these strata, which differ so much from any of the other rocks of the island, is to suppose that they were upraised at the time of the elevation of the central masses of Oahu. Whether the basalt ought to be included in this category, or ought to be viewed as a vein traversing the other beds, and standing in connection with the lavas of the interior, is doubtful. This relation would merit further examination. The only other place where I saw symptoms of stratification, was in the bounding walls of some crater-shaped ravines in the vicinity of Panalau, to the west of the promontory of Kualoa. These seemed to dip regularly outwards on all sides, from the crater, of the position, of which they gave the most indubitable evidence. Some appearances of the same nature were seen in the bounding hills of the salt lake, on the opposite side, and in some of the craters that we passed at sea, at the west extremity of the island.

There can be no doubt that the lateral ridges, descending with

such regularity from the central rides of the two chains, are ancient lava streams. This is proved not only from the nature of their component rocks, but also from the form, being narrow and more elevated on their upper part, broader and less elevated as they descend into the plains. One of the finest views of this arrangement is obtained in the streams descending from the north-east side of Mouna Raala, as seen at a distance of five to six miles from the sea-shore at Wailua. The depth of the ravines which separate these lava streams is so great, and their bounding walls, the sides of the lava streams, so steep, that the only way of reaching the central summits is by following one of these lateral ridges. Even when more open, as on the south side of the chain of Ronahuanui, they always terminate in a cul de sac or circus, bounded by tremendous precipices reaching from the bottom of the valley to the ridge of the lava stream, here, near its summit, almost elevated to the crest of the central chain, or even, as at Ronahuanui, on the side of the valley of Anuanu, to` the very highest summits themselves. It is probable that this last arrangement holds also on the north-east side of Raula, but the valley is here more elevated at its upper termination than in those which surround the base of Ronahuanui, and consequently the bounding precipices, though reaching to the highest summits, are not so lofty. These valleys are directed, as might be expected from the above mode of origin, at right angles to the central ridge, and their walls are straight, without any salient or re-entrant angles.

An interesting question suggests itself. How have the limestone cliffs in the vicinity of Kahuku attained their present position? for their summits are elevated upwards of fifty feet above the highest level reached now by the sea. They must have been under water at the time of their formation, for coral never increases above the surface, the tenants which formed it then dying. Successive elevations of a moderate amount would account very well for the terraced form assumed by the successive reefs rising above each other in regular succession. That the outer reefs become gradually deeper in proceeding out to sea, is evident from the successive increase in the magnitude of the breakers in proceeding outwards, indicating a greater depth of roll, and the outermost is invariably the largest and most magnificent.

On the Number of Comets in the Solar System; on their Light, whether Reflected or Emitted; and on the Comet which will pass its Perihelion in November 1835. By M. ARAGO.*

1. On the Number of Comets.

ALTHOUGH the question as to the number of comets has received a large share of the attention of astronomers, yet satisfactory observations are comparatively so recent, that we cannot attempt to offer more than the statement of certain probabilities on the point.

M. Lalande, in 1773, calculated that there were about three hundred in our system. On the present occasion, I purpose again to go over the ground by which he arrived at this conclusion, always, however, applying his reasoning to the numerical data which have been supplied us by the observations which have been made from the year 1800 up to the year 1830.

In this interval of thirty years, thirty-eight comets have been observed, without taking into account the appearances of the comet of 1200 days, nor that of six years and three and three quarters. This gives a supply at the rate of four comets for every three years.

If the duration of the comets which we now see was not more than 200 years, we might learn from the historians and chronologists the traces of the previous appearance of every one of them; for since the year 1600, all the celestial phenomena have been noted with sufficient accuracy. And we may here take the liberty of adding, that, whenever it was possible to observe any of these luminaries for a period of some weeks, the ellipticity of their orbits would be sensible, provided the duration of their revolution did not exceed 300 years.

Let us adopt, then, the period of 300 as the mean term for the time occupied by a comet in returning to its perihelion. Starting from any given epoch, so long as this period of 300 years has not expired, we should constantly see new comets appear; and this period once elapsed, the same luminaries would again return, though it might be in a different order.

The comets being thus all new ones during the period of 300

The above observations on Comets are taken from the third edition of Arago "Des Cometes en Generale," and the last volume of the Annuaire.

years, if every three years were to supply four comets, as has just been said, the 300 years would furnish 400 comets. Such, then, according to this mode of reasoning, would be the number of the comets of the solar system, which would be visible from the earth.

But I shall not stop now to canvass these calculations, that I may immediately address myself to considerations of a much more elevated character, by the help of which Lambert had previously endeavoured, in his ingenious Lettres Cosmologiques, to arrive at a solution of the curious problem which forms the subject of this chapter.

On the 31st December 1831, the number of comets whose orbits could be accurately calculated was 137. Let us now examine if, in their movements, these luminaries shew a tendency for any peculiar epochs, or any especial directions.

Epochs of their Passing the Perihelion.—In January there pass 14 comets, February 10, March 8, April 10, May 9, June 11, July 10, August 8, September 15, October 11, November 18, December 13; total 137.

There are evidently fewer comets in the summer than in the winter months; and it could scarcely be otherwise, on account of the shortness of the nights during the months of May, June, July, and August. The long duration of the day properly so called, and also the brightness of the twilight, cannot fail to hide from our view a certain number of these bodies.

Direction of their Movements.-Number of direct comets, 69; number of retrograde, 68; total 137.

If this comparison had been made when the number of comets whose orbits were accurately calculated was only 49, there would then have been 24 direct, and 25 retrograde, in their move

ments.

Inclinations of their Orbits.-From 0° to 10°, the number of comets is 9; 10° to 20°, 13; 20° to 30°, 10; 30° to 40°, 17; 40° to 50°, 14; 50° to 60°, 23; 60° to 70°, 17; 70° to 80°, 19; 80° to 90°, 15; total 137.

It seems to follow from this table, that the comets are more common in the greater inclinations than in the smaller. Bode had come to the same result, from considering the elements of 72 comets which were known in 1785. At the same time, we have only to glance at the catalogue of this astronomer, to per