W. W. WatU- Perlitic Structure. 19 



developed in glass. I may here state that, in a private letter to myself 

 sent just after the publication of my paper, Mr. F. Eutley called my 

 attention to this difference, insisting on the fact that whilst the 

 cracks touched one another tangentially or at a low angle in glass, 

 in quartz the angle was always a higher one. He kindly sent me 

 a beautiful perlitic rock from Jalisco, in Mexico, which illustrates this 

 admirably, and the annexed process-block, prepared from a photo- 

 graph I took of the slide, will show the feature sufficiently well. 

 However, it also draws attention to the fact that Mr. Eutley 's 

 description is a little more accurate, for this particular instance, 

 than Mr. Smeeth's, for many of the circumferential cracks do 

 anastomose icith one another in a way that cannot be explained 

 by the tangential touching of excentric circles. Indeed, so far as 

 my experience goes, the cracks in glass rarely or never reach 

 the high perfection of those in balsam. 



With regard to the second point, so far as I am able to judge 

 from his figure (loc. cit., pi. xlv, fig. 4), I do not„ observe that there 

 is much tendency in the artificial production of ' quartz perlites ' 

 for the cracks characteristic of one type of material (or method of 

 production) to invade the neighbouring material. 



In describing the Tardree rhyolite I used the term perlitic struc- 

 ture in a rather wide way to include the whole system of contraction 

 cracks which are linked up by a common mode of origin. The 

 following types of fissure were recognized: (a) polygonal cracks: 

 (b) curved cracks ; (c) radial cracks running approximately per- 

 pendicular to the polygonal fissures ; to these I would now add 

 (d) other radials joining one or two curved cracks with each other. 

 Earely only one of these types is present at a particular point of 

 a slide, usually two or three, and occasionally all four types are 

 present in one matrix perlite. 



Taking these types in order and referring to the figures and plate 

 given with my paper, 1 it will be seen that type (a) is present in quartz 

 and matrix, type (c) in the latter and sometimes in the former (pi. 

 xviii. fig. 3), and type (b) is ill developed in the matrix and absent from 

 the quartz. Fig. 5 on the same plate gives a matrix perlite with 

 curved (spherical or spiralloid) cracks (6) and radials (d) enclosed 

 in a network of polygonal cracks (a) ; all three of these types of 

 fissure traverse the quartz, two of them at least passing in from the 

 matrix. Guided by the example of the balsam one would suppose 

 that the polygonal cracks were the oldest and practically simul- 

 taneous in both constituents ; the curved ones would follow, and, 

 beginning in the glass, would cross the quartz and finish off in 

 glass. In this case the whole system seems to be bound up in the 

 two constituents. 



Fig. 6, pi. xviii is a case in which the polygon enclosing 

 a perlite is completed in quartz, and two radials at least of type (c) 

 spring from that part of the crack which crosses the quartz. Again, 

 if we agree with Mr. Smeeth that in such examples as those figured 

 in figs. 3, 5, and 6, p. 370, and in pi. xviii, figs. 1 and 2, the cracks 

 1 Op. cit., p. 371 and pi. xviii. 



