I of course utterly reprobate my wild notion of circular elevation; it is a satisfaction to me to think that I perceived there was a screw loose in the old view, and, so far, I think I was of some service to you.
Depend on it, you have for ever smashed, crushed, and abolished craters of elevation. There must be craters of engulfment, and of explosion (mere modifications of craters of eruption), but craters of denudation are the ones which have given rise to all the discussions.
Pray give my best thanks to Lady Lyell for her translation, which was as clear as daylight to me, including "leglessness."
LETTER 486. TO C. LYELL.
Down {November 20th, 1849}.
I remembered the passage in E. de B. {Elie de Beaumont} and have now re-read it. I have always and do still entirely disbelieve it; in such a wonderful case he ought to have hammered every inch of rock up to actual junction; he describes no details of junction, and if I were in your place I would absolutely dispute the fact of junction (or articulation as he oddly calls it) on such evidence. I go farther than you; I do not believe in the world there is or has been a junction between a dike and stream of lava of exact shape of either (1) or (2) Figure 2}.
(Figures 2, 3 and 4.)
If dike gave immediate origin to volcanic vent we should have craters of {an} elliptic shape {Figure 3}. I believe that when the molten rock in a dike comes near to the surface, some one two or three points will always certainly chance to afford an easier passage upward to the actual surface than along the whole line, and therefore that the dike will be connected (if the whole were bared and dissected) with the vent by a column or cone (see my elegant drawing) of lava {Figure 4}. I do not doubt that the dikes are thus indirectly connected with eruptive vents. E. de B. seems to have observed many of his T; now without he supposes the whole line of fissure or dike to have poured out lava (which implies, as above remarked, craters of an elliptic or almost linear shape) on both sides, how extraordinarily improbable it is, that there should have been in a single line of section so many intersections of points eruption; he must, I think, make his orifices of eruption almost linear or, if not so, astonishingly numerous. One must refer to what one has seen oneself: do pray, when you go home, look at the section of a minute cone of eruption at the Galapagos, page 109 (486/1. "Geological Observations on Volcanic Islands." London, 1890, page 238.), which is the most perfect natural dissection of a crater which I have ever heard of, and the drawing of which you may, I assure you, trust; here the arching over of the streams as they were poured out over the lip of the crater was evident, and are now thus seen united to the central irregular column. Again, at St. Jago I saw some horizontal sections of the bases of small craters, and the sources or feeders were circular. I really cannot entertain a doubt that E. de B. is grossly wrong, and that you are right in your view; but without most distinct evidence I will never admit that a dike joins on rectangularly to a stream of lava. Your argument about the perpendicularity of the dike strikes me as good.
The map of Etna, which I have been just looking at, looks like a sudden falling in, does it not? I am not much surprised at the linear vent in Santorin (this linear tendency ought to be difficult to a circular-crater-of-elevation-believer), I think Abich (486/2. "Geologische Beobachtungen uber die vulkanischen Erscheinungen und Bildungen in Unter- und Mittel-Italien." Braunschweig, 1841.) describes having seen the same actual thing forming within the crater of Vesuvius. In such cases what outline do you give to the upper surface of the lava in the dike connecting them? Surely it would be very irregular and would send up irregular cones or columns as in my above splendid drawing.
At the Royal on Friday, after more doubt and misgiving than I almost ever felt, I voted to recommend Forbes for Royal Medal, and that view was carried, Sedgwick taking the lead.