160 TRANSACTIONS OF THE PLYMOUTH INSTITUTION. 
Down el van on the Down itself has generally a granular-felsitic 
base, with pyramidal crystals of quartz, and hollows from which 
similar crystals have disappeared. But in the valley at Milton 
and on to Lophill it displays a much wider range. One variety 
has a compact granular-felsitic base, thickly studded with well- 
formed porphyritic crystals of quartz and felspar — much nearer 
the granitic type. Another has a semi-vitreous quartzose base, 
with porphyritic felspars and a quantity of black mica. Yet 
another is so even grained and granular that the untrained 
observer would take it for a sandstone. But the most remark- 
able illustration of the protean characters an elvan may assume 
within a short distance is supplied by the dyke cut through on 
the new Tavistock line near Shillamill. The variation here in 
breadth is so well marked that examples taken within a few 
feet of each other might be supposed to come from as many 
different dykes. The centre is porphyritic, much like the 
Grenofen, and this graduates into more even -grained and 
felspathic varieties on each side, while the outer portions are 
largely quartzose. We have the passage from all but typical 
granite into syenitic quartz-porphyry within a few yards, and 
there is every reason to believe that this, with similar elvans, 
passes into true granite in depth. 2 
Here and there in debris of Dartmoor origin, both on and off 
the Moor, are found fragments of felstone of a more or less 
rhyolitic character, remnants of genuine felsitic lava streams. 
We do not find them in situ like the granites and elvans, but 
there they lie — relics of rocks that could not have consolidated 
in depth and under pressure, and which must have had a surface 
origin. 
The best example of these in my possession is from the 
neighbourhood of Lee Moor — a grey compact felstone, partially 
banded, with well-marked fluidal structure, and enclosing fragments 
of quartz and felspar, and of igneous rock, apparently of andesitic 
type, closely resembling some in the volcanic grit from Cattedown, 
to be described hereafter. 
2 The frequent passage of granite into veins of pegmatite is worth noting 
in this connection. Many of the elvans are really micro-pegmatites, and differ 
from these veins only in the smaller size of the quartz and felspar crystals, 
between which and indeterminable felsitic matter there is every stage of 
gradation. 
