THE

GEOLOGICAL MAGAZINE.

NEW SERIES. DECADE II. VOL. V.

No. VIII.-AUGUST, 1878.

ORIGINAL ARTICLES.

I. ON THE FORM OF VOLCANOS.

By Professor JOHN MILNE, F.G.S.,

Imperial College of Engineering, Yedo, Japan.

(PLATE IX.)

ROM the short notices which are to be found about volcanos in

FROM

treatises on geology, we might often be led into the belief that they were structures so regular in contour that their form was indicative of their nature. In certain cases this appears to be so remarkably true, that I wish to add a few observations to those which have already been written upon the subject. In other cases, however, the form of a volcano does not appear to have any characteristics which distinguish it from other mountains, as will be seen by glancing through the series of views given by Humboldt of the volcanos of South America. The mountains in Iceland are also very rough. When travelling in that country, where I saw and ascended many volcanos, I do not remember that there was anything about their shape more than their steepness and general ruggedness which particularly attracted my attention. The wilderness of form presented to us by such mountains as these is so evidently the combined effect of many and varied causes, that it would be vain to seek a simple explanation for the formation of the whole.

With other volcanos which have been built up according to the formula of our text-books, that is, by the ejection and accumulation of material round a central vent, the case is different. I will endeavour to show that such mountains which, for the want of a more accurate term, have been called conical, have a particular kind of regularity which does not appear to have been hitherto noticed. Conversely, I wish also to point out, that if this particular form is observed, that from it, not only is it possible to state how the mountain has been built up, but also from any variations in this form to determine the presence and dimensions of an internal core. For these purposes I shall take the mountains which I have had the opportunity of observing since my residence in Japan. The first point to be observed about many of these mountains is the regularity 1 This, however, is a statement made from memory, for shortly after my return to England, my Icelandic sketches and note-books were unfortunately lost.

DECADE II.-VOL. V.-NO. VIII.

22

of their contour, and secondly the fact that these contours appear to be similar. Pre-eminent amongst the regularly formed volcanos of Japan with which I am acquainted I may mention Fusiyama, 12,365 feet, near Yokohama; Ganjosan, 7000 feet, near Morioka; Chokaisan, 6000 feet, between Niigata and Akita; Twakisan, 5000 feet, near Awomori; and Kumagatake, 2700 feet, near Hakodate.

If we look at the profile of these mountains near their summits, we might be inclined to call them conical; but if we look at them as a whole, as we descend we see this upper cone expanding and sweeping outwards, forming a graceful curve. The causes which might produce or subsequently affect this form may be summed up as follows.

1. The position of the crater. If this is central and remains in that position during successive eruptions, we may expect the mountain to be regular in form.

2. If the eruptions of a mountain are irregular in their action and character, the above regularity may be destroyed. Thus by paroxysmal outbursts portions of a cone might be removed,-if lava is erupted at one time and not at another, or if it accumulates on one side of the mountain more than on another, these are all causes which would seriously interfere with any regularity of contour.

3. Any outbursts on the side of a mountain, or formation of parasitic craters, will also tend in the same direction.

4. Even if the crater should be central, but the direction in which the lapilli and other materials were projected should be towards one side more than another, as I sometimes observed was the case at the eruption of Oshima in 1877, the regularity of the piling up will also be interfered with.

5. The direction of the wind during an eruption will also influence the shape of a mountain. On the Yedo side of Fusiyama the slope is less than upon the opposite side. This I believe is in part due to the wind, which at the time of the last eruption of that mountain, in 1707, was blowing in the direction of Yedo.

6. The nature, that is, the size, the specific gravity, the porosity, etc., of the materials thrown out will during the actual accumulation and subsequent consolidation give a character to the curve of the slope.

7. Lastly, we have the effects of denudation continually drawing material to lower levels and sculpturing and modifying the original contours. As this action will generally take place more or less uniformly on all sides of a mountain, it will be a cause tending not so much to destroy the regularity of a volcano as to alter the character of its slope.

When considering the various causes which may produce denudation, we must remember that there generally will have been more action near the summit of a mountain than near the base. Thus, for example, more water falls at the top of a mountain than below, and therefore the little streamlets which cause disintegration will be more active at the top of a mountain than at the bottom-the activity diminishing gradually towards the base. One of the tendencies of

such an action is to make a mountain steeper near the top than lower down, an appearance which will be again referred to.

If we wish to discover an explanation for the differences which exist between the characters of the volcanos of two countries. like Iceland and Japan, it is not to the effects of denudation that we ought to look, for denudation in the case of the volcanos of these two countries, which are either now in action or else have been so within periods which are almost historical, might, I think, be taken as being equivalent to each other. We must rather look to the nature and the products of the eruptions. In Iceland I suspect that the outbursts have been more paroxysmal, whilst the number and extent of lava flows are perhaps unrivalled. These causes have given a ragged wildness to the country and its mountains. In Japan, however, lava streams, although existing, are by no means numerous, and the mountains, which far exceed the height of those in Iceland, must, generally speaking, have grown gently upwards like a heap formed by sand falling through a funnel upon a level floor. In consequence of this it will be shown that they have in many cases assumed a form as symmetrical as any which would be expected in an engineering earthwork.

Slopes of Volcanos.-If we look at the cuttings along a railway, we shall see that their slope varies with the material through which they are run. When the material is loose and friable, the slope is more gentle than when it is tenacious and has much friction amongst its particles; for example:

For compound earth the natural slope is

50°.

40°.

38°.

22°.

16°.

When climbing up the sides of volcanos I have invariably observed that the materials resting upon any particular slope are uniform in size; and if we take a piece of lava or lapilli which is larger than the other particles, and throw it upon such a slope, we see it commence to roll rather than remain at rest. From this it appears that we ought generally to find the large pieces of material which are ejected from a volcano upon the more gentle slopes near to the base of the mountain. The reason of this evidently lies in the fact that friction depends on surface, and weight depends on volume, and that as bodies are made smaller and smaller the ratio of their surface to their volume increases more and more.

Should we be desirous of making any investigation respecting the slopes upon the face of a long escarpment, or the face of a broad chain of mountains, the theory and rules given by Rankine in his Investigations about Earthworks might be applied. In a volcano or solitary mountain, such as those which are now referred to, the conditions would be somewhat different, as we have to consider the stability of a mass of material with a variable radius, whilst in the former case the radius might be regarded as being infinitely long. As there does not appear to be any formula given in books on engineering, respecting the slope or form which would be assumed

by a heap of loose dirt, an engineering friend has shown me that it follows from Rankine's theory, notwithstanding that the same is incomplete, that the surface is that which would be produced by a simple logarithmic curve revolving about an axis,-consequently such a heap would have a slope diminishing from the top to the bottom. Looking over my note-books, amongst the steepest slopes which I have observed upon the sides of volcanos in Japan, I may mention the following: Fusiyama, 30°; Asamayama, 28°; Ganjosan, 31°; Twakisan, 30°; and Kumagatake, 40°. The first four of these are outside slopes taken near the top of the mountain. The last observation was made upon a slope of angular blocks of stone inside an old crater.

Mr. Scrope gives the slopes of volcanos as being from 20° to 35°, towards the base diminishing to 10°, and gradually ultimately to horizontality. However, whilst fully recognizing the nature of the true slope of a volcano, the sketches which illustrate works treating of this subject are, so far as can be judged by rough measurements, somewhat misleading. For example, to quote from the pictures in the works of no less a person than the great Humboldt, Umrisse von Vulkanen aus den Cordilleren von Quito und Mexico," I find that, by rough measurement, the slopes of several volcanos must be as follows: Cayambe Uren, 18,170ft., about 53°; Elcorazon, 14,820ft., 53° to 69°; Chimborazo, 17,712ft., 55°; and Ilinissa, 16,362ft., 55°. Whether these Matterhorn-like inclinations really represent reality or not, I do not know, but I should be inclined to think that artistic feeling may have caused them to have been unintentionally augmented.

Form of Volcanos.-The general form of a volcano has already been suggested whilst speaking of the inclination. To quote more fully from Mr. Scrope, he says, "It should be remarked, however, that the dykes being more numerous near the central vent, their aggregate effect in elevating these beds will be greatest there, and give them a steeper elevation near the summit than lower down the flanks of the mountain. This is one cause (but by no means the principal one) of the angle of the slope of the higher beds, and of the outer slope likewise, usually ranging from 20° to 35°; while towards the base it diminishes to 10°, and graduates ultimately to horizontality. The more influential causes of this general result are (as will shortly be shown), the frequency of lateral eruptions on the lower slopes of every volcanic mountain, loading them with parasitic cones and floods of lava, and the abundance of fragmentary matter carried down the heights by rain and floods-all combining to enlarge the base." Whilst fully agreeing with the form which is here described, the causes which are indicated as having been the principal ones in the production of such a mountain form, although no doubt, influential, are, I should think, by no means the principal ones. Of the greater number of volcanos which I have seen in Japan, the only method which I have had of obtaining their general form and curvatures has been by sketching. Of two of them, 1 Scrope, Volcanos, p. 167.

Fusiyama and Kumagatake, I have been able to obtain large photographs, of the former no less than thirty different views. As these pictures, which have a slight distortion due to perspective, were the best representation of the true form of a volcano which I could obtain, it has been assumed that they are correct, and they have been used in determining the shape of the mountain in the following manner. To investigate the form of these mountains it was first required to find an axis for them, which was attempted by drawing verticals through some objects, and their reflections, which happened to have been photographed in the foreground of one of the pictures. Afterwards it was found better to make tracings of the photographs, which were doubled until a greater portion of the profiles coincided, and then to take the crease in the paper as being the median line and axis.

Enlarged drawings made by a pantograph were also used, but these were found not to be so convenient as the tracings from the actual photographs which are represented on the accompanying sketch. In the cases which are represented, with the exception of one or two slight excrescences which are shown by shading, the profiles are coincident with a free curve. Whilst looking at these profiles it must be remembered that any apparent differences in curvature, which may be observed, are due to the fact that one profile may represent only the upper portion of a mountain, whilst another may give a view from the summit to the base. Thus, for example, profile No. I. only shows the upper portion of profile No. II. The next thing which was done was to draw a series of ordinates at 5mm apart, the length of each of which r, was accurately measured, and may be seen in the first columns of the following tables. In the second column, the sum of each successive pair of these ordinates R is given, and in the third column their differences dr.

Now it is found, as will be seen by looking at the fourth column, that is equal to a number which is nearly constant, which is the peculiarity of a Logarithmic Curve.

dr