1885.] 



Phenomenon of Crystalline Reflection. 



183 



18. The number of coloured crystals obtained by crystallisations 

 on a small scale, though very small, it is true, compared with the 

 number of colourless ones, was still so much larger than Profe,ssor 

 Mills's description of the rarity of the crystals had led me to expect, 

 that I at one time doubted whether the simply twinned crystals which 

 are so very common, if taken at a period of their growth when one 

 component is still very thin, and of suitable thickness, might not 

 possibly show the phenomenon, though the thin twin was in contact 

 on. one face only with the brother twin, the other face being in the 

 mother-liquor or in air. The circumstances of reflection and trans- 

 mission at the first surface of the twin plate must be very different 

 according as it is in contact with the brother crystal, or else with the 

 mother-liquor, or air, or some other fluid ; and yet the peculiar 

 spectrum was shown all the same whether the crystal was in air, or 

 immersed in the mother-liquor, or in rock oil- However, to make 

 sure of the matter I took a simply twinned crystal, and ground it at a 

 slight inclination to the C face till the twin plane was partly ground 

 away, thus leaving a very slender twin wedge forming part of the 

 compound crystal, and polished the ground surface. On examining 

 the reflected light with a lens, no colour was seen about the edge of 

 the wedge, where the thickness of the wedge tapered away to nothing ; 

 and that, although the' bands seen near the edge in polarised light, 

 which was subsequently analysed, showed that had colours been pro- 

 ducible in this way as they are by a thin twin stratum, they would 

 not have been too narrow to escape observation. 



In another experiment a simply twinned crystal was hollowed out 

 till the twin plane was nearly reached. The hollowing was then con- 

 tinued with the wetted finger, so as to leave a concave smooth surface, 

 the crystal being examined at short intervals in polarised light as the 

 work went on, so as to know when the twin plane was pierced. But 

 though in this case the twin plane formed a secant plane, nearly a 

 tangent plane, to the worked surface, and near the section the twin 

 portion of the crystal must have been very thin for a breadth by no 

 means infinitesimal, as was shown by examination in polarised light, 

 yet no colours were seen by reflection. I conclude therefore that the 

 production of these colours requires the twin stratum to be in contact 

 on both its faces with the brother crystal. 



19. The fact that a single bright band is what most usually pre- 

 sents itself in the spectrum of the reflected light (though sometimes 

 two or three such bands at regular intervals may be seen) seems to 

 warrant us to regard that as the kind of spectrum belonging to the 

 simplest form of twin stratum, namely, one in which there are just 

 the two twin surfaces near together. The more complicated spectra 

 seem to point to a compound interference, and to be referable to the 

 existence of more than two twin planes very near together; and in 



