BY D. MAVVSON. 469 



appears usually irregular in outline and seldom exceeding 1 mm. 

 diam. ; in two cases the form was such as to suggest hornblende* 

 as the original mineral — see photograph of one of these (PI. xxiii., 

 fig.l). The most noticeable of the minerals composing these 

 aggregates is actinolite, in perfectly idiomorphic yellowish-green 

 rods crossing and recrossing each other indiscriminately. It has 

 an extinction angle of 20*^, and the following pleochroism — 



a = light yellow. 



b = grass-green. 



C = light bluish-green. 

 n < <h> c 

 Between these actinolite laths is a base of tiny granular clear 

 felspa7\ A zone of magnetite grains is often present, forming a 

 dark ring towards the centre of the section (PI. xxiii., fig.3). 

 Zoisite in small quantity is sometimes present near the periphery. 



Addenda. — Another pebble fragment collected in the same 

 locality evidently represents a further stage in the alteration of 

 this rock. In it the original structure has almost entirely 

 disappeared, and the minerals present are all of secondary origin. 

 Zoisite and various forms of actinolite are the chief constituents. 

 The zoisite and epidote in very tiny particles form considerable 

 areas, which appear as a mosaic under crossed nicols. The 

 actinolite is fibrous, arranged in radiating and sheaf -like aggregates 

 without sharply defined borders. Pyrites is comparatively 

 abundant, scattered indiscriminately through the section. Some 

 of the lighter areas appear to represent the sites of former 

 felspar crystals, which have passed through the various stages of 

 saussuritisation, and are now represented by a mosaic of colourless 

 to faintly grey zoisite, in which are scattered aggregates of 

 actinolite-like amphibole and, occasionally, epidote in recognisable 

 grains. In the less metachemisised varieties of the rock these 

 actinolite aggregates are seen to be developed from inclusions 



* Others, again, are so irregular in form that it is thought more than 

 probable that they are the sites of original cell-spaces. 



