LITHOPHTSES IX THE OBSIDIAN OF THE EOCCHE KOSSE. 439 



steam ; but that it is associated in a most interesting manner with 

 normal spherulitic growths. Mr. Iddings ^ has himself remarked 

 that " the essential character of spherulitic growths is the crystal- 

 lization of minerals from one or more points with a radiating 

 or diverging arrangement." Spherulitic growths of this accepted 

 type undoubtedly occur in the Lipari obsidians round about and 

 within the hollows that are visible to the naked eye. But, in 

 addition to this divergent type of crystallization, convergent groups 

 of crystallites have grown inwards from the surfaces towards the 

 ceatres of the hollows. 



In the obsidian of the Eocche Rosse, the internal surface of the 

 steam-vesicles is often seen to be dusted over, as it were, with 

 minute white aggregates ; where the glass is less vesicular, these 

 white growths are dense enough to completely line the cavity, and 

 conical projections of the same materials are seen to run in towards 

 the centre, usually terminating in a spheroid of rough exterior. 

 These knotty growths may be so far developed as to bridge across 

 the cavity; and, in practically all the cases where the vesicle is 

 less than 1*5 mm. in diameter, porcellanous or clearer materials 

 nearly or completely fill it. This process of infilling has been 

 accompanied by the formation of crystalline zones in the glass 

 around the vesicles ; the outermost of these is dull grey and por- 

 cellanous, and it is often succeeded by a thin white zone, within 

 which is a distinctly fibrous layer. This growth external to the 

 original vesicle, together with the internal structures, builds up 

 lithophyses which may closely simulate normal solid spherulites. 



In microscopic sections the whole series of structures becomes 

 beautifully displayed. We have first the vesicles of an almost 

 pumiceous glass, breaking into one another, deformed, elongated, 

 and without relation to products of crystalhzation beyond the fact 

 that they thrust aside the abundant microlites that mark the lines 

 of flow (PI. XII. fig. 1). Then we find cases, in close proximity 

 to the foregoing, where brown spherulitic aggregates have begun 

 to form in the glass around the vesicle, producing for the most 

 part hemispherical forms, their bases against the wall of the cavity, 

 their convex surfaces directed towards the surrounding glass. 

 These, as viewed with crossed nieols, are true spherulitic aggregates, 

 dependent upon successive development of crystallites in the 

 glass, growth thus taking place at their convex surfaces. Yery 

 often they become continuous around the vesicle, and form the dull 

 zone, bluish -grey by reflected light, which is visible in the rock 

 with the lens or with the unaided eye (PL XII. fig. 1). In this 

 brown zone, conical or, rather, mushroom-shaped fibrous structures 

 have developed, the stalk of the ' mushroom ' projecting into the 

 cavity of the vesicle. A colourless zone of crystallites, radially 

 arranged, separates the surface of the ' mushroom ' from the 

 ordinary brown aggregation round it (PL XII. figs. 2 & 3). A 



1 'Spherulitic Crystallization/ Bull. Phil. Soc. Washington, vol. xi. (1891) 

 p. 462. 



