PRODUCTION OF VOLCANIC DYKES. 483 



in rain-channels towards the lower edge. The angular tilt of the 

 plate therefore continues to increase slowly with time, and to exceed 

 that of the slope, until at last the lava-plate stands on edge, its 

 lower part being buried in the detritus, and the upr>er part in free 

 air, the plane of the plate in this position being more or less nearly 

 transverse to the line of slope ; and this process, if continued, may 

 even cause the inversion of the bed, so that what was its upper 

 surface as it flowed from the crater over the upper part of the slope, 

 and consolidated there or further down, becomes the lower surface, 

 upon which it now reposes, and what was the bottom surface would 

 • then face the sky. But the surface of no volcanic mountain is a 

 simple cone ; its sides, at least so far as they arc deeply covered 

 with detrital material, are deeply furrowed and broken into small 

 valleys and torrent-beds, having banks of increasing steepness at 

 either side as they get nearer the bottom, and the general direction 

 being often with but little sinuosity up and down the slope, as 

 frequently seen on the flanks of Monte Vulture. Now, if a plato 

 of lava descending under the mechanism just described should reach 

 the brow of one of these rain-gullies, its course of descent directly 

 down the mountain-slope may be gradually changed, and it may 

 begin to descend by quite the same mechanism along the banks of 

 the gully in a direction diagonally transverse thereto, and may 

 finish its course by being turned up on edge as before described, 

 but now with its plane in a direction more or less nearly orthogonal 

 to that just described, and in an approximately vertical direction, 

 its plane being directed more or less nearly towards the axis of 

 the mountain. If such a plate be subsequently buried in detrital 

 matter, it will present very much the characters of a dyke of injec- 

 tion, with this difference, that the two sides of the plate will present 

 those marked differences always seen in the top and bottom surfaces 

 of consolidated plates of lava. I have observed several instances of 

 these remarkable changes of position ; thus, for example, in Au- 

 vergne, on the left and within a short distance of the carriage-road 

 from Clermont Ferrand and about five miles from that city, passing 

 over the great volcanic plateau on the direct road to Pont Gibaud, 

 on the slope of a hill of volcanic detritus, there exists a thick bed of 

 lava with many included boulders tilted thus and standing several 

 feet in the air. I observed also several such instances in 1858 

 upon the flanks of Monte Vulture, upon the side towards Melfi; 

 and I have been informed by a friend who resided in Mexico that 

 like facts are not unknown there. 



In making my lithological survey of these dykes I was prepared 

 to notice particularly whether the two opposite surfaces of the same 

 dyke were alike or different, whether the parts more or less adjacent 

 to each surface were the same as those of the interior of the dyke, 

 or differed in character or mineral composition, compactness, frac- 

 ture, &c, from the material of the latter, or whether any pyromor- 

 phic changes were observable at or near the surfaces of contact 

 where a dyke intersected another dyke or a lava bed, whether the 

 mass of the dyke was amorphous, and if jointed, in what direction 



