A PERLITIC PITCHSTONE FROM TWEED RIVER. 317 



At some instant a rupture occurs and the fissure travels round 

 with a diminishing radius of curvature as the shock travels towards 

 the centre. Hence the spiral character of the cracks. In the 

 second case, when the balsam has been highly heated the two sets 

 of cracks occur simultaneously as far as the eye can tell, which 

 may be due to the fact that the balsam is in such a condition of 

 strain that the shock caused by the formation of the polygonal 

 cracks, sends a vibration towards the centre of each interspace 

 sufficient to cause the perlitic cracking of the highly strained 

 balsam. 



In order to apply these remarks to a mass of cooling lava we 

 must seek for some agency capable of inducing the molecular 

 tension which is supplied by the friction of the ground glass in 

 the experiment. If we consider a stream of lava it is evident that 

 the surface exposed to the air and those in contact with the ground 

 become comparatively cool and hard, while the interior remains 

 still viscous. This internal portion continues to part with its 

 heat and consequently to contract. It will also experience a 

 resistance to contraction, depending for amount on the ability of 

 the external portions to withstand deformation. The result will 

 be molecular separation along certain surfaces which may — as in 

 the Canada balsam — take the form of polygonal and perlitic cracks. 

 Doubtless these only occur where there are some particular 

 relations between the strain and the viscosity, or brittleness, of 

 the medium. The various conditions necessary to induce a 

 sunicient strain and to allow of the adjustment of such relations 

 may well lead to the sparing development of the structure. I am 

 not aware that up to the present the range of perlitic structure 

 in rocks in which it occurs has been carefully studied. 



To return now to the rock which I have been describing, we 

 find there a glassy matrix full of undoubted perlites, and a number 

 of quartz grains exhibiting various curved cracks. These latter 

 are so similar in character to those figured by Mr. Watts in the 

 paper above referred to, that I may be allowed to treat them in 

 common. Apart from a general dissimilarity to the perlitic cracks 



