178 Transactions of the Royal Microscopical Society. 



worked into the line of contact, and thus attest the presence of a 

 vacuity. Of this there is not the slightest trace, the cincture of the 

 spherule being perfectly clear and glassy. Some of these glass 

 rods issue at right angles from the spherulitic bands, others 

 obliquely ; some are more or less bent, occasionally almost at a 

 right angle, as in Fig. 5. Here and there two rods may be 

 seen to anastomose and pass out into the surrounding obsidian as 

 a single rod. Fig. 7 represents a stream of granular matter 

 projected through the cortical zone of a spherulitic band into the 

 surrounding glass. Fig. 1 is a straight cylinder or elongated 

 glass rod magnified 225 diameters. It has apparently been 

 severed at two points equidistant from the ends, thus giving 

 rise to three cylindrical forms, the two terminal ones being some- 

 what thicker than that in the middle. It is, however, possible 

 that this supposition is incorrect, and that they are three elongated 

 lacuna of glass. If so, why do they lie in a perfectly straight 

 line ? a line parallel in direction with the larger spherulitic bands, 

 i.e. in the direction of the lava flow. The tension in the direction 

 of, and consequent upon, the flow of the molten matter, may even 

 have sufiiced to elongate and sever this rod. Has not this tension, 

 coupled with strain necessarily more or less rectangularly situated 

 to the direction of flow, also caused the cylindrical extrusions of 

 glass from the spherulitic bands, or do the latter phenomena 

 depend upon the contraction of the bands themselves, caused by an 

 imperfect attempt at crystallization ? Although, from the fact that 

 these extruded glass rods are often curved in opposite directions, it 

 may be assumed that they were uninfluenced by the direction of 

 the lava flow, we must also bear in mind that they pass through 

 the outer zones of the spherulitic bands in diverse directions, and 

 consequently often issued more or less obliquely into the surround- 

 ing obsidian, and equal pressure upon an overlying surface would 

 tend to curve them still more in the directions in which they 

 respectively issued. It is j)robable that their solidification was 

 very rapid. The innermost core in the spherulitic bands is usually 

 doubly refracting ; hence we may assume that devitrification was 

 set up at certain points or along certain lines, these points being 

 determined by the presence of some trifling impurity, such as a 

 minute granule or crystal of magnetite. This devitrification results 

 from the development of crystalline structure ; the crystalline 

 bands probably enveloped minute lacunfe of molten glass, and upon 

 solidification contraction of the cryptocrystalline mass ensued, the 

 outer crust of the bands became fissured, and the lacunae of glass 

 finding an outlet from the internal pressure of the band, escaped 

 through the cracks in the rod-like forms in which we now see 

 them. Finally, solidification of the magma ensued ; but it is 

 probable that these different processes took place almost syn- 



