98 



Messrs. D. Herman and F. Rutley. 



tions, where they shoot into the unaltered glass, cannot be clearly 

 made ont. In most cases their terminations appear to be rounded, 

 while in others they have a rectangular aspect, suggestive of a basal 

 plane or an edge normal to the principal axis. The groups are not 

 separated by prismatic joint planes, but the divergent crystals of 

 adjacent groups appear to slightly overlap. The directions of extinc- 

 tion indicate that they are possibly rhombic forms. When magnified 

 about 570 diameters the individual crystals seem frequently to consist 

 of linear aggregates of minute globulites, but this appearance is 

 possibly deceptive, and in some cases the crystals exhibit no such 

 structure. The terminations of the crystals pass rather irregularly 

 into the adjacent glass, giving the edges of the crystalline groups a 

 fringed aspect somewhat like the pile of velvet. The adjacent glass 

 shows colourless spheroidal specks or globulites. 



Section E 2 is taken parallel to one of the faces of the six-sided 

 prism, and is, indeed, a shaving of the de vitrified crust. Mr. Cattell 

 succeeded in making a section the full size of the face, and from this 

 the drawing, fig. 1, Plate 2, was made. In this drawing a basal and 

 lateral edge are shown, and it will be seen that from these edges 

 divergent groups of crystals pass inwards. With the exception of 

 this fringe, which represents more or less oblique sections of the 

 crystalline groups which constitute the devitrified crust, the re- 

 mainder of the face shows only a polygonal network, the polygons 

 being the cross sections of prisms. It has, in fact, the same structure 

 as the margin, only the crystalline groups are in this part cut trans- 

 versely to the direction of their growth, while at the margin they are 

 cut obliquely, for the section being taken a little distance inwards 

 from the surface of the face trenches slightly upon the crystalline 

 groups of the adjacent faces, both lateral and basal. The section as 

 originally cut was so feebly translucent that an endeavour was made 

 to reduce its thickness. This, however, resulted in its almost total 

 disintegration along irregular cracks without materially increasing- 

 its translucency. 



Specimen F. This is a completely devitrified sphere of light- 

 coloured bottle-glass 18 mm. in diameter, devitrified in two operations 

 under the same conditions as Specimens I, K, and D. Under the 

 microscope a section taken through the centre of the sphere shows a 

 somewhat irregular circumference, which is accounted for when the 

 surface of the original specimen is carefully examined, for it is seen 

 to be pitted with numerous small cavities, and to have a rough fritted 

 and imperfectly glazed aspect. The irregularities of this surface are 

 due to the impressions of sand- grains, a few of which may still be 

 detected adhering to the surface. The glass has evidently undergone 

 incipient fusion, and the crystallisation in the immediate neighbour- 

 hood of the sand-grains is very small and confused. This irregularly 



