214 MR. J. PAKKINSON ON THE HOLLOW SPHERULITES [May 19OI, 



(ii) Cavities with a definite form. — Almond-shaped cavities, 

 frequently elongated till they resemble a rift in the rock rather 

 than an ordinary vesicle, are common. Such a rock resembles a 

 series of plates, 1 inch thick and upwards, imperfectly welded 

 together, and containing many interspaces. Usually these plates 

 are lithoidal, but contain patches and streaks of the black glass. 

 For instance, we find an elongated porous spherulite an inch or 

 so in length, contained between two hard compact flow-bands and 

 embedded in glass at either end. It consists of an aggregate of 

 crystals which in places are only loosely set together, and at one 

 end contains three small hollows, the greater length of which is at 

 right angles to the direction of flow. Such a spherulite, when the 

 cavities are more fully developed, presents an arrangement of con- 

 centric rings lying in a plane parallel to the flow-bands. This is 

 the typical lithophysal structure.^ The concentric rings, which in a 

 cross-section bridge a cavity, consist of a rather coarse granular 

 aggregate of crystals of felspar and tridymite (fig. 2). From the 



Fig. 2. — HoTloiv spherulites. 



,Jn%^^^^^4>^^ 





'"■'\ \ n* " ' :'^l'vri>V"vK"A'v^V-%>:J^Vr... 



III 



II 



I = Outer edge of one of the series of concentric rings which distinguish 

 the lithophysee of Obsidian Cliff, showing the projecting crystals of 

 tridymite (t) and felspar (/). x40. ■» 



II = Interspace between two rings of a lithophysa from near Wrockwardine, 

 showing remnants of projecting crystals (see PL VIII, fig. 2). x45. 

 Ill = Remnants of a mineral resembling tridymite, from near Wrockwardine. 

 Seep. 221. x45. 



almond-shaped rifts we obtain gradations to other kinds of hollow 

 spherulites. Thus, by decreasing the length of the cavity parallel 

 to the plane of flow, and broadening it in a direction at right 

 angles to that plane, we arrive at an ordinary hollow hemispherical 

 spherulite lying on a flow-band. Such a cavity shows relation to 



1 See Iddings, 7th Ann. Rep. U.S. Geol. Smw. (1885-86) p. 264 & pis. xii 

 ,& xiii ; for chemical analyses of the obsidian and lithophys£e, see ibid. p. 291. 



