518 MESSES. P. LAKE AND S. H. KEYNOLDS ON THE [Aug. 1896^ 



stretches west towards Maes-angharad. This being so, the most 

 striking feature of the whole, as shown upon the map, is the 

 contraction of the outcrop where the streams east and west of the 

 town have cut their valleys deep down into the igneous mass. On 

 the high ground the width of the diabase is great, and the boundaries 

 follow the contour-lines, reducing, however, the prominence of the 

 curves. Where, on the other hand, the surface is lower, as in the 

 valleys of the streams, the outcrop is narrow, and the boundaries- 

 are straight and independent of the contours. These facts indicate 

 that at the level of the higher ground the diabase spreads out into 

 a broad mass, the under surface of which, near its boundaries, is 

 gently inclined ; while at the level to which the valleys have been 

 carved it forms a comparatively narrow vertical-sided dyke. The 

 transverse section in fact is like a section of a mushroom. 



Near Maes-angharad the mass narrows again ; and this no doubt 

 is partly owing to the lower level of the ground in that neighbour- 

 hood. But it may also be due in part to the Derwas fault. The 

 downthrow of this is on its south-eastern side, and if it faults the 

 diabase as well as the sedimentary deposits, its effect will be to 

 bring to the surface on its north-western side a lower and therefore 

 narrower part of the mass — to expose, in fact, the stalk instead of 

 the head of the mushroom. 



Close to Dolgelly there is a smaller patch of diabase which forms 

 the slopes immediately above the town, and this is perhaps a part 

 of the northern expansion of the great mass, from which, however, 

 it has been separated by denudation. The southern boundary of 

 this patch is a curve, the convexity of which points up the slope on 

 which the rock occurs. The surface, therefore, which separates the 

 igneous rock from the sedimentary deposits near it, is not vertical, 

 but dips down hill. Moreover, the bed of the Afon Aran and 

 the right bank of the stream above Pandy'r-odyn do not show 

 diabase, but ash in the lower parts of their courses. These streams 

 have cut through the diabase, and have exposed the beds below, 

 and the diabase is therefore a sheet upon the hillside, no doubt 

 with an irregular base. 



In its mushroom-like form the great mass of diabase which 

 stretches across the Dolgelly fault is not unlike a laccolite, but in 

 its relations to the beds below it offers certain striking peculiarities. 

 Its northern boundary west of the fault runs nearly parallel to 

 the lava-band already described, so that here the mass appears to 

 rest upon the surface of the bed below it. But along its southern 

 boundary, whereas the surface of separation between the diabase 

 and the stratified beds below it is shown by the map to be nearly 

 horizontal, the stratified beds themselves dip at a fairly high angle 

 to the south. The diabase, therefore, no longer rests upon the 

 surface of the beds, but upon their edges. Moreover, as we have 

 already seen, a considerable thickness of slates which should occur 

 between the lava of the Bryn-y-gwin Wood and the ashes of Graig 

 is invisible. These slates are probably concealed below the diabase. 



On the eastern side of the fault the beds of ash, etc., which lie 



