516 Scientific Proceedings, Royal Dublin Society. 



granules of magnetite occur in addition to the plagioclase. The 

 groundmass and the dusky spherulites are full of skeleton-crystals 

 resembling those in slags, the long axes of which are often curved ; 

 they are arranged here and there in sheaves, but without relation 

 to the subsequent spherulitic crystallization, except that small ones 

 are thickly massed upon the margins of the spherulites (PL xxi. 

 fig. 3). While the spherulites show patches of low colours of the 

 first order between crossed nicols, radial structure is barely trace- 

 able. 



With a power magnifying 400 diameters, the dark skeleton- 

 crystals are seen to have long spindle-shaped axial bodies, which 

 are built up of globulites, so that they remind one of the spicules 

 of Alcyonidse. Cross-bars are set on many of these, apparently at 

 right angles, and often in two series perpendicular to one another 

 and to the axis. When very thin, their colour is a warm brown ; 

 but suspicious stainings around some of them make one believe 

 that they were originally skeletons of magnetite, as in many 

 tachylytes, ^ and that they have become coloured by further oxi- 

 dation. A few felspathic microlites occur, with forked termi- 

 nations to the prism ;^ but the other prevailing individualized 

 constituents consist of long and often pale-green rods, which have 

 developed quite independently of the magnetite, traversing the 

 skeleton crystals at all angles. They are sometimes very long 

 (•5 mm.) in comparison with the axes of the magnetite ; I regard 

 the coloured ones as microlites of pyroxene. These coloured rods 

 are often gathered together in long groups, which are without 

 visible axes, the individual microlites lying parallel to one another 

 and perpendicular to the long axis of the group ; it is as if one 

 were to cut the string of a bundle of firewood, spread out the 

 sticks in a parallel series, retaining their original orientation, and 

 then remove the string. For convenience we may term this 

 arrangement the "palisade-structure" (PI. xxi. fig. 4). 



The central mass of the dyke yields the most slag-like section 

 of a natural product that I know. The crosses of magnetite, more 

 or less developed, have the same orientation over areas 2 mm. 



1 E.g. the rocks of Screpidale and Some Point, Quart. Journ. Geol. Soc, 

 vol. xxxix., pi. XIV., figs. 3 and 4. 



2 " Crenulites " of Kutley ; Min. Mag., vol. ix., p. 267. 



