Microscopic Characters of Devitrified Glas, 



91 



ments which, run approximately at right angles to the surfaces which 

 constitute the boundaries of the devitrified specimen. In polarised 

 light the general aspect of the section is peculiar, and strikingly 

 resembles a patchwork rug made of the skins of tabby cats. Further 

 on we shall endeavour to account for this brindled appearance, which 

 is represented in fig. 1, Plate I, as seen between crossed Nicols. The 

 vertical edge of the section seen on the right of the field is the trace 

 of one of the parallel faces of the devitrified plate-glass. The 

 N.TV. portion represents part of one of the terminal triangular 

 areas, while the remainder shows some of the other crystalline allot- 

 ments. 



Specimen "No. 115, Section B. This is a particularly interesting 

 section. It is in fact one of the green surfaces of the devitrified 

 plate-glass, i.e., present surface, ~ inch of glass having been ground 

 away, and we can easily trace in it the polygonal structure already 

 alluded to. Between crossed ISTicols the polygonal areas are sharply 

 defined and are irregularly clouded with crystalline aggregates, which 

 appear dark. On rotating the section through 90° these dark 

 aggregates become light, while the previously light portions become 

 dark ; we are, in fact, looking on the ends of bundles of crystalline 

 rods. These polygonal areas are the cross sections of fasiculi of 

 divergent crystals, and the boundaries of these polygons are shrink- 

 age cracks, giving rise to a columnar structure, while the columns, 

 like those of basaltic lavas, have their longest axes normal to the 

 cooling surface, rig. 2, Plate 1, shows the general appearance of this 

 section, magnified eighteen linear, between crossed JSTicols. As it seemed 

 possible that greater amplification might afford more information 

 concerning the nature of the little crystals which constitute these 

 bundles, a ^-inch objective was used, with the result shown in fig. 3, 

 Plate 1. Only dark hazily-defined spots could be discerned between 

 crossed Nicols, which became light on revolution of the section, while 

 previously light portions became dark. The section has, in fact, the 

 appearance of what is known as crypto-cry&talline structure, and 

 resembles, to a certain extent, some of the felstones, which, from other 

 evidence, are known to have been once vitreous lavas. An examina- 

 tion of this section proves then that the polygons are the cross 

 sections of the crystalline allotments of Section A, and -that those 

 allotments are longitudinal sections of polygonal, often pentagonal, 

 prisms. Whether or not the polygonal jointing is connected with the 

 crystalline developments, which it sheaths and separates, is a matter 

 open to discussion. The Section B, when held between the eye and 

 candle flame, presents the illusive appearance of being studded with 

 concavities or convexities, from which it, we think, may be inferred 

 that the radiate arrangement of the crystalline fasiculi originates at 

 or about the centre of each polygonal area on the original surface of 



