Microscopic Characters of Devitrified Glass. 107 



specimen is a slab of J -inch. British plate, about 4 inches by 3 inches 

 in diameter, and upon one of its surfaces a straight cut or scratch, 

 about 2 inches long, was made by a diamond, producing an exceed- 

 ingly fine crack, extending at the edge of the plate to a depth, of over 

 1 inch in a direction approximately normal to the surface upon which 

 the scratch was made, and gradually dying out to the end of the 

 diamond cut. The crack was sufficiently fine to show Newton's rings. 

 The specimen was then completely devitrified by heating continuously 

 for nine days afc a bright red heat, a temperature considerably 

 higher than was employed in the case of Specimen 126, 127, and 

 it was subsequently cracked across in the direction of the line 

 marked BJB in fig. 1, Plate 4. Fig. 2 represents the fractured 

 surface. At each end are the usual triangular areas, formed by lines 

 of arrest, but the line of arrest which usually joins the apices of 

 these triangular areas is here interrupted by another series of crystal- 

 lisations which have emanated from the crack produced by the 

 diamond scratch. In the lialf of the plate nearest to the scratched 

 surface we have, indeed, a reproduction of what has taken place at 

 the outer edges of the plate, the result being a nearly equilateral 

 triangular area of crystallisation, bisected by the crack already men- 

 tioned. This crack, however, passes a little beyond the median line 

 of arrest, and from its termination the crystallisation radiates and 

 ends against a curved arrest line, as shown in fig. 2, Plate 4. 



That devitrification does not always proceed in the orderly and 

 uniform manner seen in Specimens 115, 147, and, indeed, in nearly 

 all of the examples already described in this paper, will be best 

 realised by reference to the figures of Specimens Nos. 132 and 136 on 

 Plate 4, figs. 4 and 5. Fig. 3 in the same plate, Specimen No. 137, is 

 a small slab of glass partially devitrified. The crust has been formed in 

 the usual way by crystallisation proceeding from the surface inwards, 

 but the process has been arrested, and where the outer crnst is broken 

 away a core of somewhat cracked but perfectly clear glass is seen, in 

 which no spherules or other crystalline bodies are visible. In Speci- 

 mens 132 and 136, however, the result has been different, for after 

 a slight external crust has been formed, devitrification has also 

 started from numerous points within the glass, giving rise to a well- 

 marked spherulitic structure. 



Why these spherules have been formed instead of a gradually 

 increasing crust is a matter which we hope to explain in a subsequent 

 paper. 



