66 Scientific Proceedings, Royal Dublin Society. 



was common over the surface of the polygonal form : oleavage was 

 generally parallel to the ends of the rectangle in the rectangular 

 forms. 



On first approaching the subjeot, I formed the hypothesis that 

 some of the symmetrically extinguishing rectangular forms were 

 beryl cut parallel to axis of prism. The absence of hexagonal 

 sections threw doubts on that hypothesis. Some of the other 

 rectangular forms I thought were orthoclase developed on the 

 zone pti (Levy). Such a zone shows large base, reotangular, 

 with extinctions parallel to sides ; orthopinacoid large, with extinc- 

 tions parallel to sides ; clinopinacoid small, with extinctions paral- 

 lel to sides in "orthose non-deforme" at 5° in "orthose deformed 

 But this zone should show, when cut squarely, the axial angle of 

 63° 33 '. It was never found thus in the sections. The colours 

 of polarization, too, were not those of orthoclase. 



It was evidently possible, also, to account for the appearances 

 by supposing the crystals orthorhombic, in which case, further, 

 both forms might be supposed to be different views of one and the 

 same crystal. 



Now the angles of the polygon were all about 150° when 

 twelve-sided ; and in eight-sided figures, not uncommon, one set of 

 faces produced intersected at 90°, another at 60° and 120°. To 

 what orthorhombic mineral were such angles to be assigned ? Not 

 being acquainted with any such, I had set the mineral for the most 

 part down as "doubtful," with the suggestion that some of the 

 forms might be orthoclase, when I had recourse to an apparatus I 

 devised about this time for investigating the melting points of 

 small fragments of minerals. This apparatus is briefly described 

 in Nature (vol. xxxiii., p. 15), where I call it a " meldometer," or 

 measurer of melting points. By the help of this apparatus I 

 differentiated them from every substance I had ever suspected as 

 being present. I must explain, however, how I succeeded in 

 obtaining the crystals isolated and free of the matrix. 



I mentioned before the cavities eaten by decomposition in the 

 large hexagonal crystals of mixed beryl and orthoclase. These 

 cavities appeared filled in part with a rusty-brown powder, and in 

 part with a frail skeleton of hard matter (felspar) clinging to the 

 walls or loose in the cavity. On removing this debris, crushing 



