366 PEor. J. w. jrDD ox the teetiart a^d 



molecules ; and all the appearances presented by the lamellar 

 twinning of these felspars seem to me to be much more in harmony 

 with the view that it is so induced than with the supposition tha!t 

 the structure is an original one. 



"With respect to the period at which these twin-lamellae were 

 developed in the crystals of j)lagioelase felspar, there is every reason 

 to believe that during the cooling-down of the crystals from the 

 high temperature at which they were developed, many such struc- 

 tures would arise ; for they are found alike in the felspars of arti- 

 ficially-formed rocks and in those of drusy cavities and open veins. 

 But the more carefully the positions and relations of these twin- 

 lamellae are studied, the more clearly does it become apparent that 

 external causes, such as the pressure of adjoining crystals and the 

 other accidents to which they have been subjected since their first 

 formation, have played no inconsiderable part in bringing about the 

 final result. In PL X. fig. 1, a crystal of felspar is represented, 

 which admirably illustrates these conclusions. 



[At the time when this paper was read I was not aware of the 

 important series of experiments which had been carried on by Herr 

 Forstner, and which are described in the Zeitschr. fiir Krystall. &c., 

 vol. ix. (1884) p. 333. Forstner took plates of the untwinned 

 plagioclase felspar of the rock from Pantellaria which he has de- 

 scribed, and by heating them to various temperatures and cooling 

 them again, he succeeded in developing the twin-lamellae, and even 

 in producing reentering angles on the edges of the plates similar to 

 those which have been artificially produced in fragments of calcite. 

 These experiments prove conclusively that the twin-lamellae of 

 plagioclase felspar can he developed by the application of mechanical 

 force ; and that they are constantly so produced in rock-masses, I 

 have in the foregoing pages given reasons for concluding.] 



The Pyroxenes. — There has been a very great conflict of opinion 

 concerning the pyroxenes in the gabbros of the Western Isles of 

 Scotland. ITaccuUoch, while rightly identifying the pyroxene in 

 the great mass of these rocks as augite, stated that hypersthene 

 occurs in the Cuchullin Hills of Skye, in Ardnamurchan, and at one 

 point in the island of Eum *. In these statements he was perfectly 

 correct, but in going farther and describing the rock-masses of the 

 Cuchullin Hills and the western parts of Ardnamurchan as 

 hypersthene-rock, Macculloch fell into an error ; for in all these 

 rocks, as was pointed out by Zirkel, the predominant pyroxene is 

 diallage, the foliated form of augite, and not hypersthene. We 

 shall show, however, that augite frequently occurs in these rocks in 

 an altered form, which so curiously mimics the colour, lustre, and 

 general appearance of hypersthene that Macculloch's error was a 

 very natural and pardonable one. 



Descloiseaux's researches in 1862 showed that the pyroxenes can 

 be divided into two groups, those in which lime is the predominant 

 base, crystallizing in the monoclinic system — the augites — and those 



* A Description of the Western Isles of Scotland, vol. i. pp. 385, 390, 

 413, 503. 



