MELAPHYKES AND FELSITKS OF CARADOC. 541 



point the strike is »S. 80° W. and the dip about 55° towards the 

 north. A microscopic section of this rock, however, shows scarcely 

 any traces of lluxion-banding. 



Irregularly-shaped vesicles, Ulled with quartz &c., are often 

 present in those rhyolites, and they occasionally show a tendency 

 towards elongation in a definite direction. This is well scon in 

 some of the exposures of rhy elite occurring at the top of Caradoc 

 <;lose to and within the Camp ; but the forms of these vesicles, when 

 viewed under the microscope, are often remjirkably irregular, 

 throwing out processes in all directions, the latter frequently con- 

 stricted to mere threads near their points of origin from the main 

 vesicle, then ex])anding and finally tapering to sharp points. In 

 spite of this irregularity, however, they may be seen to have a 

 rudely linear arrangement in the rock. Fluxion structure, as 

 evidenced by streams of microliths, is not to be detected in rocks 

 which have undergone such complete devitrification, save perhaps in 

 hazy banding, produced by slight differences in crystalline texture. 



For the pur])ose of ascertaining how far certain structures in vitre- 

 ous rocks may be destroyed or rendered invisible through subsequent 

 alteration, I have examined a number of sections of unaltered ob- 

 sidians and pitchstoncs in search of perlitic and fluxion structures, 

 so delicate that devitrification would almost infallibly obliterate 

 them or render them so indistinct that they could no longer be 

 recognized with any certainty. The following are a few notes on 

 the subject, wliich may possibly be of some interest: — 



1. In a perlitic obsidian (from Schemnitz, Hungary) the breadth 

 of the perlitic fissures ranges from about -ygVcr ^^ 2tV"o ^^ ^^ inch. 

 Such a continuous perlitic fissure, when seen in section, may often 

 be observed to thin away from the larger dimension to nothing. 

 One can easily imagine that the devitrification of such a rock would 

 result in the total obliteration of the more delicate fissures. 



The fiuxion-bauding in this section consists of streams of micro- 

 liths. A portion of one of the broadest lines in one of these streams 

 measures -pr^-,;) of an inch. Devitrification might not obliterate such 

 a band, but it might easily render the recognition of its component 

 microliths impossible. 



2. In a section of a Mexican obsidian no fluxion-banding what- 

 ever is visible under the microscope, the only indication of flow 

 consisting in the uniform direction of elongation of included gas- 

 pores. The numerous microliths present in the section lie with 

 their longest axes in all directions. 



There are, therefore, in such a rock no structures which would 

 bear testimony to its origin after devitrification. The result of such 

 change would be simply a felsite without fluxion-banding and with- 

 out perlitic structure. 



In cases such as these, mode of occurrence and associations in 

 the field could alone give a clue to the original nature of the rock. 



The best example of fluxion structure which I have yet met with 

 in the Caradoc district is in a rhyolite-tufi" occurring at Bowdler's 

 Chair at the southern extremity of the (jiacrstones ridge. The 



