A CONTRIBUTION TO THE PETROGRAPHY OF BENGUELLA. 553 



high content of soda and titania. It is zonal, with hour-glass structure, faint pleo- 

 chroism, and strong dispersion of the bisectrices. In some slides the' smaller 

 crystals, and the borders of the larger ones, have a greenish tinge, indicating some 

 admixture of the segirine molecule. Apatite occurs abundantly in large dusty 

 crystals with rounded corners. In some, an addition of fresh material has grown 

 around them with sharp crystalline boundaries, in a manner which recalls the quartz 

 crystals of the Penrith sandstone. A little sphene occurs in the usual wedge-shaped 

 crystals ; and there is frequently a border of granular sphene fringing irregular masses 

 of titaniferous iron-ore. 



The affinities of this rock are clearly with the type described as shonkinite by 

 Pirsson. In his latest work Rosenbusch * defines shonkinite as a hypidiomorphic 

 granular rock of the alkaline series, characterised by the mineral combination potash- 

 felspar and nepheline, and by the predominance of coloured constituents. This 

 definition would make it the melanocratic facies of the nepheline-syenite group. 

 Iddings t defines the rock as a phanerite characterised by alkali-felspar and equal or 

 nearly equal amounts of mafic minerals, with small but notable amounts of fels- 

 pathoids such as nepheline or sodalite ; and he treats it as the mesocratic facies of 

 nepheline-syenites. In the original shonkinite of Square Butte, Montana, nepheline, 

 along with sodalite, was regarded as a rare accessory mineral, and was, indeed, only 

 detected by chemical means. J The rock was thus regarded as the mesocratic or 

 melanocratic facies of alkali-syenites. The later work of Pirsson showed, however, 

 that his type shonkinite contained 5 per cent, of nepheline and 1 per cent, of sodalite. § 

 The term shonkinite, in fact, has been used to cover rocks which belong to the meso- 

 cratic and melanocratic divisions of both alkali-syenites and nepheline-syenites. 

 Recognising this, Lacroix has proposed to restrict the term to the melanocratic types 

 of the alkali-syenite family, || and to use another term for the equivalent rocks of the 

 nepheline-syenite family.H In his latest description Rosenbusch himself speaks of 

 " the typical, nepheline-poor shonkinite of Montana." ## The Ochilesa rock falls 

 most naturally under the original definition of the term. 



The quantitative mineral composition of the Ochilesa shonkinite has been 

 determined by the Rosiwal method, with the result shown in Table V, 1. It is 



* Gesteinlehre, 3rd ed., 1910, p. 199. t Igneous Rocks, vol. ii, 1913, p. 230. 



| Weed and Pirsson, Bull. Geol. Soc. America, vol. vi, 1895, p. 415. 



§ Pirsson, " Geol. and Petr. of Higliwood Mountains, Mont.," U.S. Geol. Surv., Bull. 237, 1905, p. 104. 



|| Lacroix, Nouv. Arch, du Mus., 1902, p. 179. 



IT Lacroix proposes to use the term theralite in this sense. This, however, ignores Rosenbusch's plain intention 

 of using this term for the plagioclase-nepheline rocks of pin tonic habit. It is true that the original " theralite " of the 

 Crazy Mountains, Montana, on which Rosenbusch erected the group, was erroneously described by Wolff as carrying 

 plagioclase, and is in fact a shonkinite in the sense of Rosenbusch and Iddings. Lacroix's usage of the term thera- 

 lite, however, merely perpetuates the original mistake in the diagnosis of the Crazy Mountain rocks. When 

 Rosenbusch discovered the mistake, he promptly withdrew these rocks from the theralite group, and placed them 

 under shonkinite. He then created as his type theralite the unquestionable plagioclase-nepheline rock of Duppau in 

 Bohemia, thus preserving his term theralite with its original connotation. 



** Loc. cit., p. 203. 

 TRANS. ROY. SOC. EDIN., VOL. LI, PART III (NO. 14). 80 



