A PERLITIC PITCHSTONE FROM TWEED RIVER. 315 



In a mass of lava on the other hand there is no such directive 

 influence at work, so that the perlites would lie with their axes in 

 all directions, and if they are surfaces similar to the ideal one, we 

 should get the sections exhibited in tigs. 3 to 6. As a matter of 

 fact we do get figures closely resembling some of these in sections 

 of perlitic rocks, but the majority of the curves in the natural 

 structure are considerably different, and we are led to inquire 

 what modifications of the ideal surface is likely to occur in a 

 natural glass, and whether such modification will yield curves 

 similar to those most frequently observed. The most obvious factor 

 of disturbance will be a want of homogeneousness, which will I 

 presume, be admitted to be a feature of natural glassy magmas. 



The probable effect of this factor will be to produce a less 

 symmetrical surface, and one in which the different turns — -instead 

 of being all rolled on one axis — are rolled on different axes. (The 

 axis being the line joining the points of contact of any turn with 

 the surrounding turn). For instance, if the axes of the turns in 

 fig. 4, be displaced in the plane of the paper, we get curves of the 

 type shown in fig. 7. If they are displaced in other directions as 

 well, curves similar to figs. 8, 9, 10, must result with considerable 

 variations in detail. To these considerations we must add the fact 

 that the curves are frequently segmental or intermittent, and that 

 the different segments may represent portions of more than one 

 surface. These last set of curves bear a striking resemblance to 

 those which may be observed in sections of perlitic rocks. Types 

 approaching the previous more regular curves may also be observed 

 but not frequently. In confirmation of these remarks I refer to 

 the photograph of perlitic cracks (Plate 45, fig. 3), and to the 

 drawings of actual perlites (Figs. 11 to 15). 



If these suggestions as to the nature of perlitic structure are 

 admitted to fairly represent its true character, I may be allowed 

 to define it as follows : — Perlites are cracks of more or less irregu- 

 larly spiralloid character, occurring in the interspaces between 

 sets of polygonal cracks ; in section these appear as rudely con- 

 centric curves which are frequently in contact at various points. 



