520 ProceedinrjH of the Royal Irish Academy. 



From these observations it is clear that "we are dealing with 

 crystals in which the opposite faces are not parallel ! This anomaly 

 explains a difficulty encountered in measuring the zone [001], due to 

 the impossibility of exactly adjusting the crystals, so that the bright 

 signal should always fall on the cross wires of the telescope ; if the 

 adjustment were correct for any five faces it would invariably fail for 

 the sixth. The deviation, though not large, was obvious. The 

 measarements in this zone gave the following : — 







I. 



II. 



010 : 



: 110 



62° 



62° 35' 



110 ; 



; 110 



55 55' 



56 1 



no : 



: OfO 



61 24 



61 41 



OlO ; 



; 110 



! 118 41 



62 20 



110 : 



; 110 



) 



55 29 



110 : 



; 010 



62 



61 54 





360° 



360° 



These observations sufficiently illustrate the irregularities by which 

 the crystals are deformed ; if they are inconsistent with the symmetry 

 of the oblique system they are no less so with that of the anorthic 

 system, for even centro-symmetry is absent. 



A microscopic examination shows striation on the prism faces, 

 whence the train of images that are seen on measiu'ement ; and on the 

 basal plane a pavement of polygonal areas, some of which project 

 beyond the others, with an evident tilt. Small as the crystals are 

 they are far from simple, and may be be best regarded as crystal com- 

 plexes, simulating and making a close approximation to a simple 

 crystal form. It might be suggested that unequal contraction or 

 cooling may have helped to produce a deformation, for the attached 

 extremity of the crystals is immersed and rooted, as it were, among the 

 other constituents of the rock, and these extend into its substance for 

 some distance beyond the general surface of attachment. Fizeau gives 

 for the mean coefficient of expansion of hornblende 0*00000866, of 

 orthoclase 0'00000753, and of quartz 0-000011 ; and it is easy to cal- 

 culate from these data what kind of effect the suggested action would 

 produce. If we suppose the case of a cube consisting for half its 

 height of a mixture of orthoclase and hornblende, and for the remaining 

 half of pure hornblende, it is easy to show that the unequal contraction 



