the vessel, by the subtraction of heat, may cause its external surface to contract, and this may mechanically contract it internally, and so heat may be evolved. This explanation is nearly the converse of the previous one of the steam which has been given. There, a chemical expansion first happens, followed by a mechanical expansion, by which heat is involved. Here, in the metal, a chemical contraction first happens, succeeded by a mechanical contraction, by which heat is evolved.

On the Mode of Formation of Crystalline Limestone, Contact Products, Crystalline Silicide-Slates, and Unstratified Crystalline Silicide-Rocks; with Preliminary Observations on the present state of Geology, and on the Methods of Investigation pursued in that science. By B. M. KEILHAU, Professor of Geology in the University of Christiania. Communicated by the Author.*

It would be a happy state of things if, by means of direct observation, we could every where acquire a perfect knowledge of the structure of that portion of our globe which consists of the known mountain-rocks, and if we could obtain a elear view of the part performed by each mountain-rock, or generally by each mineral mass, as an architectural element, in a construction so complicated. We should thus create an indestructible basis for geology, and could provide for it a foundation of data really belonging to itself, which would confer on it an independence that is at present awanting. It is an acknowledged truth, that we cannot be very doubtful as to the origin of a mountain-rock when the relations of its masses are elearly placed before us; at least, in such a case, altogether erroneous hypotheses would scarcely be possible. If, for example, we see that an entirely irregular mass lies between sedimentary strata, and branches out into these, we cannot, in refer

*This literal translation of Professor Keilhau's manuscript has had the advantage of his revision.-EDIT.

fence to its mode of formation, place it in the same class with the strata; and if we see it bounded on all sides by the sedimentary rocks, we cannot attribute to it an eruptive origin; in short, by means of this simple but correct mode of examining the subject, both the hypotheses which are most frequently brought forward would be rejected. In the same manner undoubtedly, every hastily adopted opinion as to the origin of rocks would, by a complete knowledge of the facts of the structure of our globe, be at once reduced to its true value. But this is not all, for this perfect, this, so to speak, anatomical knowledge, would also afford many positive contributions to an insight into the actual nature of the object,—to an understanding of the origin and cause of the phenomena presented.

It can admit of no doubt that geology would undergo very essential changes, were we able in such a manner to have an unobstructed view of the structure of the earth. So long tas, owing to the impossibility, in most cases, of obtaining Ja knowledge of facts by direct examination alone, observers always endeavour, in a greater or less degree, to assist their observations by supposing, by means of considerations as to causes and actions, how the phenomena under investigation must be, this great evil must result, that the science does "not obtain possession of perfectly unmixed data. What are termed facts frequently very little deserve the appellation. Such data are partly the work of the understanding of the 'observer, and, as they are more or less the result of a precon'ceived explanation, the science becomes in the same degree illusory, being founded on such a basis; nay, the whole procedure remains a mere movement in a circle, inasmuch as the "observer explains what the explanatory reasoning itself has just silently advanced. This great evil could not exist in the case previously supposed. Imaginary relations would then never take the place of those really observed, partly because there would be no necessity for it, and partly because what was inaccurate would be much too easily pointed out, and in this way the mania for making hypotheses would be suppressed. On the contrary, the opportunity afforded of furnishing the science with that pure foundation of unfalsified data which it now wants, would be eagerly embraced, and the first endeavour

of every investigator would be directed to the distinct exposition of the really existing phenomena. When this was entirely finished, but not sooner, the observer would proceed to the task of treating of the causes of the geological phenomena, and of their mutual connection. In such labours, therefore, the actually existing geological relations themselves would always constitute, as the very nature of the subject requires, the first and most important point, and that not merely because these are precisely the object of investigation, but also because they themselves can afford such essential aid in the elucidation of their true nature. In the next place, what we learn by the observation of geological processes going forward before our eyes would become available; and, thirdly, then, for the first time, would those elucidations come into consideration, which can be borrowed from other sciences, viz. from chemis-try, natural philosophy, and astronomy.

We know sufficiently well that this natural order of things is at present not the existing one. As we cannot obtain such irrefragable knowledge of geological facts as would be possible on the supposition made above, geologists are so far from assigning to results obtained by geognostical examination the just rank in their theories, that they often consider themselves authorized to modify the exposition of geological phenomena, in order to accommodate it to the explanation created out of those sciences which only stand in more or less remote connection with geology. It is sometimes even the case that they go still farther, and from time to time believe themselves justified in putting observation entirely out of sight, and supplanting it by ideal phenomena according to views derived from the most varied sources, which, however, are beyond the proper territory of geology. In this manner, the proper order is exactly reversed; the science comes to rest on a foreign basis; and that which ought to be the result of pure geognostical observation becomes, at least in a greater or less degree, a mere construction, for which opinions derived principally from chemistry have solely furnished the materials.

Thus, if geological relations were freely and openly exposed -to observation, so that no difference of opinion could be possible, with regard to the real nature of the facts; if, on this

account, nothing were borrowed from ideas derived from other sources; and if no one were to replace the real by imaginary phenomena, then, undoubtedly, in theorizing, these relations would with justice attain to the first rank under all circumstances a state of matters which does not at all exist at the present day. Besides this principal change which, upon our supposition, would take place in the science, it may be worth while to mention another, which would also be very important.

Although, under the circumstances supposed, much that is now problematical would be cleared up, yet, on the other hand, the very possession of geognostical data which could not be impugned, and which could admit of no accommodation, would very often cause us to meet with inexplicable phenomena. This, however, would not be at all remarkable, and still less would it be a stumbling-block. That the human mind, which in no instance is able to understand nature to the very foundation, should here find an exception, would occur to no one; and we should here likewise have to encounter darkness beyond the nearer or farther limits which explanation could reach. It is worthy of attention, that at the present time matters are in a very different position; of phenomena which cannot be explained we now hear very little in geology! precisely in that science in which so much must be obscure, it seems as though every thing were perfectly understood. The method adopted goes the length of requiring that every phenomenon shall be placed in such a light that its cause can be ascertained, otherwise no attention will be paid to it, or its description will be regarded as inaccurate. To the uninitiated this must appear in the highest degree absurd. When, indeed, we reflect how much is still obscure respecting the origin of mineral bodies, we must be astonished at such a state of matters in geology, and such a mode of proceeding among geologists. But in this, we only see a direct consequence of the existing eircumstances. Instead of geological phenomena, which are adduced as facts, being considered as quite certain by and for themselves, and only by and for themselves, we cannot now, as it is so difficult to observe with perfect accuracy, have full confidence in the apprehension of the observed facts, until we find that they har

monize with what, according to our theory, we assume ought to exist in the case under consideration. In order to be able to believe, we must also be able to understand. As the objects cannot here be made palpable, we are not in such a situation that belief must exist under all circumstances, whether we understand or not. It is even considered as a very correct principle of investigation, when the question regards any geognostical appearance, not to recognise it as a pure fact until counsel is taken from sciences to which in reality only a secondary voice should be assigned in such matters. Thus it has become quite usual in geology not to tolerate phenomena which seem to be in discordance with the present position of the sciences alluded to. These are rejected on the pretext either that they are imperfectly observed, or that they constitute isolated abnormal phenomena, to which no weight can be given. An ample field is thus thrown open for caprice, and the science is exposed to this disadvantage, that a multitude of important facts are not introduced into its archives, facts belonging precisely to that class on which might be founded principles, that are not at all, or at least not easily, to be obtained from any other science but geology itself. As the key to such facts cannot be found in those other sciences, these very facts are thrown aside but this is done to the irreparable detriment of geology; for, as already remarked, it may be the case that it is only by the study of these facts themselves, and their analogies, that the most important information can be obtained. I hope afterwards to illustrate this more fully by distinct examples.

What I have already said, is sufficient to shew that geology is by no means in the best possible situation. However, inasmuch as the conditions for a more desirable position, such as that mentioned at first, with which the actual one was compared, depend on an impossibility, of what use can it be, it may be asked, to think of a change of that kind? But on considering the matter more attentively, it will be found that very much can be done in this respect. We have it completely in our power to approach much more nearly to such a position than we are at present; it is possible in very many cases at least to substitute direct observations for ideas. VOL. XXXVI. NO. LXXII.-APRIL 1844.

Z