in Labrador Felspar. 327 



same mineral to similar cavities, though we are no longer able to 

 see their individual outline, or to recognise them in any other 

 way but by their united influence. 



The coloured parallelograms above described are, in general, 

 parallel to the face P, Fig. 3, of the primitive form, as given by 

 HAUY ; and in no specimen which has come under my examina- 

 tion, have I ever found them coincident with the plane of the 

 common changeable colours which have been so long admired in 

 Labrador Felspar. The largest generally occupy one plane; 

 but I have found another set arranged in another plane, while 

 others have their reflecting edges in a variety of different posi- 

 tions. This effect will be understood from Fig. 4. which repre- 

 sents the images reflected from all the different colorific planes 

 in a specimen in my possession. When we look into the speci- 

 men, we see the image C of the candle formed by the ordinary 

 polished surface cut at random. Let the felspar be now turned 

 round till AC, the line joining the candle C, and the great mass 

 of changeable colour A is in the plane of reflection, A being seen 

 by rays incident at a greater angle than C. When this is done, 

 we shall see a series of nearly coincident coloured images of the 

 candle at D, which are the reflections from the parallelograms 

 shewn in Fig. 1 . At E, there is another, set of nearly coincident 

 images, fainter and less coloured than those at D. At B there 

 is a third set, but they are still fainter and more indefinite. 

 Through these three sets of images there passes an arch of red- 

 dish-brown light, extending on each side towards F and G, and 

 formed by minute needle-shaped cavities, which being nearly of 

 equal diameter in every direction except their length, must re- 

 flect light in whatever direction it is incident, provided it fall 

 nearly in a plane perpendicular to their length. 



We come now to the examination of the changeable colours 

 of the spar, which, so far as I know, have never been submitted 



