470 PEOF. W. S. BOULTON OX THE [NOV. I904, 



passes into peroxide, alkalies derived from the action of sea-water enter into 

 combination, the quantity of alumina remaining almost constant.' 



In the palagonite-tuff of Pontesford most of the iron is in the 

 peroxide-state, while the percentage of lime is high, and magnesia 

 low. It must be remembered that the rock coDtains a fair amount 

 of secondary calcite, so that both original lime and magnesia may 

 have been removed from the glass during the formation of pala- 

 gonite. It would be difficult, or impossible, however, to demonstrate 

 these changes in a tuff with secondary deposits (calcite and zeolites) 

 in the matrix, as is the case with this Pontesford specimen. 



(c) Andesite-Lavas. — The andesite-lava, which, as already 

 noted, occurs as lapilli in the tuffs, is also found interbedded with the 

 tuffs, though covering a relatively-small area. It has been found very 

 difficult, in the field, to separate some of these andesite-lavas from the 

 newer basic rocks, for both may be fine-grained, with a dull blue- or 

 grey -greenish colour. Typically, however, the andesite is bluish- 

 green, weathering yellowish-green, line-grained, with small white, and 

 often squat-shaped, felspars, easily recognized with a lens, and small 

 soft black specks, giving greenish scratches, which under the micro- 

 scope are found to be portions of the interstitial glassy matrix con- 

 verted into palagonite, together with vesicles filled with a chloritic 

 substance ; while the compact varieties of the intrusive dolerite 

 usually weather reddish-brown, and the felspars are less prominent. 

 Where the andesite occurs near to the intrusive dolerite (and indeed 

 it is never far from it) it is often darker than usual, owing, as 

 microscopic examination shows, to the development of large numbers 

 of minute magnetite- or ilmenite-grains, so that it becomes increas- 

 ingly difficult to distinguish it from the finer dolerite. Moreover, 

 there is always the possibility of the dolerite showing through among 

 the andesite-lavas, as, indeed, it actually does in one or two places 

 (516 d). The difficulty of separating these rocks does not completely 

 disappear when a microscopic examination is made : for, as will 

 appear in the following descriptions, many of the mineralogical and 

 structural characters are common to some specimens in both groups. 

 By slicing a large number of rocks at all the doubtful points, and 

 repeatedly noting their field-relations in the light of the knowledge 

 obtained from an examination of these rock-sections, it has been 

 found possible to distinguish the two groups, and map their 

 boundaries with tolerable accuracy. 



At the south-western end of the gulley, at the foot of the northern 

 face and opposite ' Agglomerate- Crag,' is a typical specimen of 

 the andesite-lava (57), interbedded with green-and-yellow finely- 

 laminated halleflinta ; and a few yards to the west (516) is a group 

 of crags showing two similar thin beds of lava (516 a, 516 c), a few 

 yards wide, separated by relatively-soft green palagonite-tuff (516 b), 

 while a small sill of intrusive dolerite abuts against 516 c. The 

 same lavas are seen in among the hallefiintas and palagonite-tuffs 

 all along the lower slope of this part of the hill, and up to the line 

 of the intrusive dolerite (525, etc.). 



