Kyanite in the mainland Lewisian complex

Abstract

Numerous P–T estimates (listed in Table 1) have been produced in recent years for the granulite-facies Badcallian metamorphism of the Lewisian complex on the mainland of NW Scotland. A check can be made on these estimates by identifying the stable Al2SiO5 polymorph(s) found in rocks interleaved with those for which the estimates are given. Newton and Perkins (1982) suggest that sillimanite is the stable Badcallian polymorph throughout the Lewisian (presumably following Davies 1974), and this is indeed consistent with their P–T estimates. We have, however, been unable to confirm the presence of sillimanite in the area studied by Davies (1974), although quantities of cummingtonite, closely resembling sillimanite at outcrop, are widespread. Moreover, this area is one of intense deformation and recrystallisation during Inverian and Laxfordian events, and the mineralogy of these rocks reflects the amphibolite-facies conditions prevailing during these later events, rather than Badcallian granulite-facies conditions. Kyanite has been reported at two localities north of Loch Maree (Carnmore and Meall na Meine) by Peach et al. (1903) and Crane (1978) respectively. The kyanite in these occurrences is considered by Crane (1978) to be of Inverian or Laxfordian age according to the relationship between mineral growth and deformation. Both kyxanite and sillimanite occur in paragneisses on South Harris (Dearnley 1959; Coward et al. 1969) but geochronological investigation of associated rocks (Cliff et al. 1983) indicates that the granulite-facies assemblages in these rocks developed after 2.2 Ga, much later than those on the mainland. There is one well-documented occurrence of kyanite . . .