54 PEOP. J. W. JUDD OlsT TEKTIAET 



similar composition in other districts, such as the Auvergne, Bohemia, 

 or Italy, we shall find that there are groups of characters in the 

 rocks of each of these areas which are highly distinctive. 



These facts point to two important conclusions. The first is 

 that there are distinct petrograjjJiical provinces, within which the 

 rocks erupted during any particular geological period present 

 certain well-marked peculiarities in mineralogical composition and 

 microscopical structure, serving at once to distinguish them from 

 the rocks belonging to the same general group, which were simul- 

 taneously erupted in other petrographical provinces. The second 

 conclusion is that Antrim, the Inner Hebrides, the Faroe Isles, and 

 Iceland were, during the Tertiary period, included in the same petro- 

 graphical previa ce*. 



This petrographical province was one of vast extent. Prom Iceland, 

 which covers an area of nearly 40,000 square miles, the Faroe Isles 

 lie 250 miles to the south-east ; while the St. Kilda group, an extreme 

 outlier of the British Archipelago, lies another 250 miles still further 

 to the south. Although I have not been able to visit the last-men- 

 tioned remote group of islands myself, the clear descriptions of Dr. 

 MaccuUoch leave no doubt as to the remarkable identity in the 

 character of its rocks with those of Skye, Ardnamurchan, and Mull. 

 Mr. Alexander Ross, of Inverness, who has visited the islands, 

 rendered me the greatest service by kindly supplying me with an 

 interesting series of photographs and specimens. The former show 

 that the characteristic modes of weathering of the acid and basic 

 rocks so admirably exhibited in the Inner Hebrides are there exactly 

 repeated ; and the latter enable me to prove that the gabbros, 

 dolerites, basalts, quartz-diorites, and granites of St. Kilda present 

 precisely the same features as are found in the corresponding rocks 

 of Mull and the other volcanic centres. 



More than 100 miles to the south-east of the St. Kilda group, we 

 find another, and perhaps the largest, centre of eruption in the dis- 

 trict, the portions of the mass of erupted matter which have escaped 

 destruction by denudation, forming the greater part of Skye, liaasay, 

 and the adjoining islands. The Shiant Isles constitute perhaps only 

 an outlying member of this group. 



South of this centre of eruption, at a distance of less than 20 

 miles, is a third, the materials thrown out from which constitute 

 the gronp known as the Small Isles (Eum, Cauna, Eigg, and Muck). 



* I have elsewhere pointed out (' Volcanoes : what they are and what they 

 teach' (1880), p. 202) the family hkeness which is found to exist between the 

 different classes of lava erupted within a certain area during a given geological 

 period, and their marked distinctness from those of other areas or other periods. 

 JSTowhe-re do we find tJiis more admirably exemplified than in Bohemia and 

 Hungary, where volcanic outbu.rsts were taking place during a great part of the 

 Tertiary period. While in the former district phonolites and tephrites, with 

 nepheline- and leucite-basalts, were being erupted^ in the latter none of these 

 lavas made their appearance, their place being taken by andesites and quartz- 

 andesites, rhyolites and felspar-basalts. In precisely the same way the several 

 varieties of the igneous rocks of the Brito-Icelandic petrographical province 

 are seen to exhibit the most striking resemblances to one another, while they 

 show very well-marked differences from those of other areas. 



