when the sun is in the tropics, or, perhaps more correctly, when the earth is in its perihelion or aphelion. It is a minimum near the equinoxes, or when the earth is at its mean distance from the sun. The greatest daily variation is least in the winter, and greatest in the summer. The greatest difference of the annual variation is 0,0359.

Suppose I call the horizontal part of the magnetic force =H, the whole force = F, the magnetic inclination n; then is F=Hx cos n. Hence it follows, that H may be variable, although F be constant, allowing only the inclination to be variable. From a series of observations which I made with an inclinatorium (a dipping-needle) of Dollond's, I found, that the inclination during the summer is about 15' greater than what it is in the winter, and in the forenoon about 4 or 5' greater than what it is in the afternoon; which agrees perfectly well with the former observations.

I next made an experiment like that of October 1. continued without interruption from noon till noon, during an aurora borealis, which clearly proves the influence of this phenomenon of weakening the magnetic force, and consequently shews the relation between magnetism and electricity." Similar experiments were made by the Professor, which prove that the magnetic intensity is always weakened when the moon crosses the equator.

In the course of last year, Professor Hansteen spent a few days in Copenhagen, and resided in the Round Tower. Here he repeated his experiments, and found, to his surprise, that the duration of 300 vibrations was not less than 836",57, although he found it in a garden below only 779. Now, from Bugge's Observationes Astronomica, it is known, that the walls of this tower, which is 126 feet high, are 4 feet 4 inches thick; and another hollow cylinder of 4 feet 6 inches inner diameter is contained within it, round which a walk of seven turnings leads to the top. After several experiments below, which all gave 787" for the mean duration, Hansteen returned to the tower, and found in the highest part of the tower the duration of 300 oscillations=842",37 1 turning below,

2 turnings lower, got

11⁄2 turning lower,

2 turnings lower,

Down below, within doors,

836,57

837,30

834,43

804,07

813,00

n

S

This attracted the attention of Professor Hansteen, and, after having further extended his observations on this point, he obtained the general result, that at the foot N of any vertical object SN, the cylinder oscillates quicker in a, to the north end thereof, and slower in b to the southward thereof; but on the contrary, at the upper part S, it oscillates quicker on the south side at d, and slower on the north side at c. Hence, he draws the conclusion, that every vertical object, of whatever material it is composed, has a magnetic south pole above, and a north pole below.

n

N

b

a

ABT. X.-On the Ancient Volcanoes of Auvergne. By CHARLES DAUBENY, M. D. M. G. S. and Fellow of Magdelane College, Oxford. In a Letter to Professor JAMESON.

DEAR SIR,

IN my former communication, I took notice of the Volcanic Rocks found in the immediate neighbourhood of Clermont, all of which, with the exception of a few mountain caps, composed of basalt, which were mentioned in the close of my letter, appeared to me of post-diluvian origin. I shall next proceed to another class of rocks, which, although probably belonging to the same general head, seems, nevertheless, to present several important differences.

The Mont d'Or range of Hills, which is the principal subject of the present paper, must, I believe, be referred exclusively to the class of Ancient Volcanoes, inasmuch as they are found, like the rocks on which they lie, at equal levels, on either side of the valleys which intersect them; and, therefore, if we suppose that they were once in a state of fluidity, a point which I hope afterwards to be able to establish, evidently prior to the formation of the latter.

These are the rocks which appear to me to supply the connecting link between what is called the Newest Flotz-Trap Formation of Werner, and the products of existing volcanoes; for those who have taken the trouble of perusing my former letter, will perceive that the rocks therein described differ in so many important points from the former, that we should hardly be justified in arguing from the one to the other. Should it, however, appear that there is reason for concluding the chain of Mount d'Or to be of volcanic origin, I know not how we can help extending the same inferences to many of the basalts of our own and other countries, to which it seems to present such striking and numerous analogies.

Two classes of volcanic rocks compose the chain of Mount d'Or. On the summit of the great Table Land of which it consists, is a basaltic formation, associated with a sort of trap tuff or breccia*, and other rocks.

,

Underneath this is a porphyritic felspar-rock seen exposed in the valleys, which has generally obtained the name of Trachyte.

Below all is seen the granite, which seems to constitute the original substratum throughout the whole of this Province; the volcanic, and even the fresh water formation, hardly deserving to be considered as parts of the crust of the earth, but being rather in some sense extraneous to its composition +.

We shall consider these rocks in succession, beginning with the highest in the series.

1. The Basaltic Formation of Mont d'Or comprises several rocks, differing from each other much in appearance and external characters.

The rock which has induced me to give this name to the whole, is a compact and sonorous basalt, generally containing

Although in general very averse to alterations in nomenclature, I cannot help agreeing with Dr MacCulloch in the propriety of substituting the term "TrapBreccia" for Trap-tuff, as the latter word is objectionable, not merely as inexpressive of the true character of the rock, but as conveying an erroneous idea of it.

+ M. d'Aubuisson, in his late work on Geognosy, not unaptly compares the Primitive Rocks to the skeleton of an animal, the Secondary to the muscular parts, and the Volcanic to the parasitical animals that are produced upon its sur. face.'

disseminated crystals of hornblende, having frequently an acicular form, and sometimes a few of olivine. Associated with it is a vesicular rock, to which the name of Lava may not improperly be given, since the cavities which it contains possess no appearance of having once been filled, like those in the amygdaloids of basaltic countries, and its general aspect obliges us to refer it to the same class with the undisputed products of volcanoes now in activity. We meet also with scoriæ, either in detached fragments, or with portions of compact trap, the whole cemented together by iron-clay, so as to constitute a species of volcanic tuff. I found the order of superposition on a hill which I examined near Lake Gery, a few miles from the village called "Les Bains de Mount d'Or," to be as follows:

On the summit a thin bed of scoriæ. Underneath, a tuff, containing fragments of the more compact, united with the vesicular variety of lava, but in some places in a state of such extreme division, that the whole might be mistaken at a distance for red sandstone. Beneath all was a very compact and crystalline basalt, made up of a confused assemblage of these acicular crystals of hornblende, crossing each other in all directions, intermixed with felspar.

It would seem that this is the general order of superposition; and the pressure of the vesicular matter above may possibly have contributed to give to the rock underneath the hardness and consistency which it is found to possess.

Thus, we may observe in many of the recent volcanic products near Clermont, that the masses which lie underneath are less vesicular than those near the surface, although I have noticed in my former communication one fact which seems to militate against the universality of this remark. It is also true, that I have found basalt among the mountains of Auvergne, uncovered by any more porous rock; but is it not probable, that the same causes which excavated the valleys, and removed all traces of the craters, which we must suppose to have once existed, should have in these instances swept away the upper strata, which, from their looser and more vesicular structure, were less able to oppose resistance? May not this account for the occurrence of clinkstone, the least decomposable of trap-rocks, chiefly on the summit of hills; for this position seems hardly

referable to any law in the original formation of the rocks of which it forms a part, since we now and then meet with it underneath other rocks *.

From the description we have given, it will be easily understood, that the subjacent rock, the trachyte, is only seen exposed on the sides and in the bottom of the valleys, the great and elevated Table Land, which composes the range of Mont d'Or, and extends with little interruption into Cantal, having its upper strata composed of basalt, and the other rocks associated with it; the scorified matter appearing to diminish in quantity, in proportion as we recede from the loftiest part of the chain. That these strata are all volcanic, is pretty convincingly proved by the vesicular structure of those which lie uppermost; for it seems impossible to assign a different origin to the basalt, and to the scorified matter superimposed *; yet it is difficult to form an opinion in what direction this immense stream of lava can have flown, or from what point it could have been ejected. I noticed, indeed, one or two lakes, which gave me much the idea of their having once been craters, the Lake Servieres in particular, to the north of the village of Les Bains, the sides of which are composed like the Puy Pariou, near Clermont, of vesicular lava, and its form equally round and regular. Near it is a little hillock, not exceeding 50 feet in height, of a conical shape, consisting chiefly of the same materials.

Yet, if this was a crater, the situation of the country must have greatly altered since the period at which it was in activity, its height, although absolutely great, being relatively to that of the rocks surrounding it, very inconsiderable. It lies, indeed, in the midst of this great Table Land, if it may be so termed, little, if at all, elevated above the general level.

2. The Trachytic Formation is essentially composed of crystals of glassy felspar, imbedded in a base of the same materials. Its fracture is more commonly rough and earthy, but is not unfrequently compact. In the latter state it is, that augite, mica,

As in the Isle of Lamlash underneath sandstone.

+ M. d'Aubuisson has remarked at Pradelles, near Clermont, basalt incumbent on a bed of scoria.