1)1 



the colourless variety may be irregular or determined by accidents, such as 

 the contiguity of two hornblende crystals. (1) Here we appear to have 

 evidence that the colourless hornblende is entirely of secondary origin, and 

 that it has been deposited on original hornblende exactly as the secondary 

 quartz has been deposited on the grains of that mineral in many quartzites ; that 

 is in optical continuity with the original crystal or crystalline grain. That 

 colourless hornblende arises in connection with the serpentinisation of oliviiie 

 is proved by the phenomena described by BECKE ; and we appear to have 

 illustrations of the same feature in many of the British picrites. It seems 

 advisable to say " in connection with " rather than " as a result of," because 

 it is probable that the phenomenon is dependent on simultaneous changes in 

 other minerals. 



The brown hornblende is characteristic of the least altered rocks and 

 is probably original, for it sometimes shows definite external forms in which 

 not only the faces of the prism but also those of the clinopinacoid (010) are 

 developed. 



Cross sections of the prism in which the cleavage planes intersect at 

 angles of about 124 and 56 change in colour, as the stage is rotated (2) above 

 the polariser, from a rich brown (often a clove-brown) when the short 

 diameter of the iiicol bisects the acute angles formed by the intersections of 

 the cleavage cracks, to a pale brown when the stage is rotated through 90. 

 Such sections arc approximately parallel to the greatest and mean axes of 

 elasticity ( and ft), the former bisecting the obtuse, and the latter the acute 

 angles of the parallelograms formed by the cleavage cracks. The colour for 

 rays vibrating parallel to is therefore pale brown, and that for rays vibrating 

 parallel to ft, brown or rich clove brown. Orthopinacoidal sections may be 

 distinguished by parallel cleavage cracks and straight extinction. They 

 contain the mean and least axes of elasticity, and show a feeble pleochroism, 

 or else remain unchanged as the stage is rotated over the polariser. 



a, pale brown. 



ft and 7, deep brown. 



In the green varieties of hornblende which often occur in the same 

 rock, 



a, pale brown or green, 

 ft and 7, bluish green. 



It is interesting to note that the difference between the green and the 

 brown hornblende in these rocks is most marked when the sections are 

 viewed with rays vibrating parallel (3) to the ft and 7 axes. Crystals which 

 present the most striking differences of colour when viewed under these 



(1) See BOXNEY. Q.J.G.S.. Vol. XLI,, Plate XVI., Fig. 2. 



(2) In testing for pleochroism the stage and riot the polariser should always be rotated. 

 Indeed, for ordinary petrographical work, it is advisable to keep the polariser always fixed in 

 one position, and to remember the direction of the short axis. 



(3) The rays of polarised light are assumed to be due to vibrations parallel to the short 

 axis of the nicol. 



