Yol. 54-] OF BOULA.T BAY (jersey). 115 



parts the original flow-lines are generally found to persist. The 

 latter are not sufficiently pronounced to prevent the secondary 

 structure from being formed. 



By incident light the unaltered part of the fragment appears 

 brick-red, as is so commonly the case in these rocks ; the radialized 

 part, on the contrary, is a dull greenish grey, with golden-yellow 

 polarization-tints. 



Yi(T, 5. Diagram of fragment showing growth normal to its sides, the 



direction of ivhich is indicated by the three arrows. 



X24. 



[The space enclosed by the inner triangle still shows the original flow-lines, the 

 direction of which is indicated by the innermost arrow. The space external 

 to the inner triangle is now occupied by the secondary growth.] 



In some cases the fibrous growth, starting at right angles to the 

 bounding-walls, has left no trace of the original, and the appearance 

 is then less striking than when, proceeding uniformly inward, it has 

 stopped short of effacement of the original flow-lines, leaving a 

 central part in which these are still clearly visible, contained by 

 lines parallel to, and equidistant from, those defining the fragment. 



It is impossible to avoid the conclusion that this radial structure 

 has arisen from the secondary heating-up to which the fragments 

 have been subjected, and in this respect it is closely analogous to 

 that described by Prof. Bonney from Arran.^ 



Prof. Cole,^ in a paper based on a specimen of rhyolite brecciated 

 by flow from Rocche Rosse, has described exactly the same phe- 

 nomenon. He gives it as his opinion that ' spherulitic crusts will arise 

 where fragments of obsidian, sufficiently consolidated for fracture, are 

 subjected to renewed heating, or even to the slow passage of the 

 residual heat from the lava-flow of which they form a part.' 



1 Geol. Mag. 1877, p. 506. 



2 Min. Mag. vol. ix (1891) p. 273. 



I 2 



