NOTES AND EXHIBITS. 1081 



and multiple-twinned. The sides of the microscopic crystals 

 parallel to the principal axis are sharply defined while the ends 

 are frayed out. They are arranged tangentially round the large 

 grains of olivine. From this disposition, and from the absence of 

 evidence of corrosion, it would appear that they were formed at or 

 near the surface of the volcanic orifice, and were not derived from 

 great depths. The olivine grains on the other hand appear to be 

 " erratics," like the large augite crystals, as may be inferred from 

 their rounded outlines, the corrosion they have undergone, which 

 in some cases has allowed glassy material from the base to 

 ])enetrate to the heart of the crystal, and from their size as 

 compared with the microscopic crystals, some of them measuring 

 one-twelfth of an inch in longest diameter. A tew of the olivine 

 crystals are traversed by cracks along which greenish or reddish- 

 brown decomposition-products have formed. The majority how- 

 ever are remarkably free from alteration and show clear boundaries. 

 The edges, however, when examined under a ^ inch power, are 

 seen to be systematically corroded or fused along certain parallel 

 lines (probably the solution planes of the ciystal) so that the 

 edges appear to be ribbed like a file, or toothed like a saw. For a 

 short distance from the edges of the olivine crystals the glassy 

 base has become devitrified possibly owing to the highly heated 

 olivine grains keeping the temperature of the glass surrounding 

 and interpenetrating them at a higher temperature than the rest 

 of the glass, and so allowing it to cool more slowly, which would 

 of course favour its devitrification near the point of contact. One 

 of the larger olivine grains is penetrated by a steam-pore, or 

 minute tube, the inside of which is partly coated with partially 

 devitrified glass, which has assumed arborescent forms ; and in 

 another part of the same crystal is another minute steam-pore 

 coated with black dendrites, probably dusty magnetite which 

 formed simultaneously with the dendritic magnetite in some of the 

 cracks in the glass during the cooling of the lava. The pore 

 extends from the centre of the olivine grain to its outer surface 

 where it empties into another pore, which in turn passes into a 

 line of globulites in the glass. A very significant fact is the 



