OCCURRENCE IN ANCIENT BRITISH LAVAS. 165 



splienilitic origin. There may be made out, however, among later 

 lavas two classes of large spherules, which are merely extensions of 

 structures familiar enough upon an ordinary scale. 



We know, for instance, how in many obsidians the spherulites 

 prove on microscopic examination to be mere cloudy aggregations, 

 with perhaps the faintest trace of radial structure. Thus in a 

 beautiful example from Iceland, numerous brown irregular patches 

 are seen separating out along the lines of banding; since crystal- 

 lization has not gone so far as to condense into microliths the 

 globulites of which they are formed, these little nodules show the 

 banding and even the fluidal structure of the matrix passing unin- 

 terru])tedly through them. Their relation, indeed, to these pro- 

 bably earlier structures is that of concretions in sedimentary rocks 

 to the beds in which they lie (PI. lY. fig. 1). 



In proportion as slower cooling allows of the radial grouping of 

 the constituents, so the drawing together of the material towards a 

 centre tends to sharpen the limit between the spheruHte and the 

 glass, until a crack is often the result. Usually, however, the outer 

 border of even these well-marked forms contains a large amount of 

 glass, and frequently concentric coats of similar cumulitic and glassy 

 matter intervene between the radial zones. 



Taking now sections of the obsidian wdth '' Lithophysen " from 

 Beaver Lake, we find in places no distinct boundary between the 

 grey nodular aggregations and the banded structure of the glass 

 (PI. lY. fig. 2). The lines of minute spherulites run into the larger 

 felsitic masses ; and the latter, though probably somewhat altered 

 since the consolidation of the lava, contain throughout the same rod- 

 like form of crystallites as is abundant in the surrounding matrix. 

 The felsitic substance is distinctly cryptocrystailine, its ill-defined 

 particles polarizing in colour ; and a very rough attempt at concentric 

 arrangement occurs in portions of the nodules, the lines being brought 

 out in hand-specimens by unequal weathering. Such irregularly 

 bounded nodules may be broadly regarded as segregations of rhyolite 

 in rhyolite-glass. They are localizations of such stony material as, 

 in longer process of cooling, might have finally spread throughout 

 the mass. 



The equally irregular central hollows bear no resemblance to 

 steam-vesicles, while signs of alteration are very evident in the 

 felsitic matter round them. Flakes of haematite, red by transmitted 

 light and giving an iron-reaction before the blowpipe, are developed 

 in the powdery mass ; and the beautiful rod-like crystallites, clear 

 and colourless in the matrix, have, in these decomposing nodules, 

 been either converted into or replaced by a red-brown ferruginous 

 product. While the specific gravity of the glass is 2*41, that of the 

 nodule, obtained by means of the bottle, is only 2*31, a result 

 pointing towards hydration of the constituents rather than towards 

 such a concentration of silica as Szabo refers to in the case of the 

 Hungarian " Lithophysen " *. 



If, then, the nodular obsidian of Beaver Lake appears to reproduce 

 * Jahrb. der k.-¥. geol. Reichs. 1866, p. 89. 



