[47] Vol. i., chap. 8. This traveller did not visit the cave, but quotes from Olafsson and Pállsson, p. 927.

[48] This interesting letter was brought to the author’s notice by Dr Attilio Hortis, Director of the Bibliotheca Civica, Trieste. This young and ardent scholar has published for the centenary festival of Petrarch (June 1874), certain political documents hitherto unprinted; they prove Petrarch to have been, like almost all the great Italian poets, a far-seeing statesman in theory if not in practice.

[49] Bochart (in Chanaan, i. 40), quoting Diogenes and Dercyllides of Tyre, whose tables, according to Photius (loc. cit.), were dug up by order of Alexander the Great, explains Thule to mean in Phœnician “tenebrarum insula.” But this etymology reminds us of the Semitic origin applied to Britain.

[50] The Icel. is Thilir, men of Thela-mörk, mark of the Thilir, the Norwegian country now called Thilemarken.

[51] Dr Charnock remarks that “Thule” is the name of a river in Glamorganshire, of a place in Silesia, and a town in Westphalia; also that “Southern Thule” was a title given to a part of Sandwich Island, the southernmost region discovered by Captain Cook in January 1775. Lt. Wilford’s Pandit invented a Pushkara Dwipa under the Arctic circle, corresponding with modern Iceland. Camden (Britannia) warns us, not unnecessarily, against confounding the “insula in ultimis et extremis Borealis Oceani secessibus longè sub Arctico Polo,” with the Indian “Tylis” or “Tylos” (Bahrayn?), of which St Augustine (lib. xxi. 5, De Civit. Dei) says, “Tylen Indiæ insulam eo preferri cæteris terris, quod omnis arbor quæ in eâ gignitur nunquam nudatur tegmine foliorum,” doubtless alluding to the palm. Strabo, we believe, does not mention “Tylos;” Pliny refers to it in three places (Nat. Hist., vi. 32, and xii. 21 and 22).

[52] To which may be added, neglecting the “Automata” of classical and mediæval times (Pliny, i. 89; Ruspe, de Novis Insulis, etc.), Arons Island (1628); Sorea of the Moluccas (1693); the offsets of Santorin (1707); Stromöe (1783); Graham Island, near Sicily, which, in 1831, was thrown up to a height of 750 feet, and the three outliers of Santorin (1866). These little worlds enable us to study Earth in the art of parturition.

[53] From Palagonia in Sicily, where it was first described (1838) by that savant (see pp. 222-483, and 802, Dana’s System of Mineralogy, Trübner, London, 1871). The specific gravity is 2·43, and the fracture mostly conchoidal. The distinguished chemist, Professor Bunsen (Sect. ix., § 1), who, succeeding in producing artificial Palagonite, gives it iron, either magnetic or peroxide, and “some alkali,” a vague term: Dr W. Lauder Lindsay adds minor constituents, felspar, augite (hornblende), jasper, olivine, obsidian, hornstone, chalcedony, and zeolite. Professor Tyndall (Royal Institution, June 3, 1853) offers the following table:

Oxide of iron,36·75
Alumina,25·50
Lime,20·25
Magnesia,11·39 (not found by Dr Murray Thomson).
Soda,3·44
Potash,2·67
100·00

In 1872, only a single and a very poor specimen of this highly interesting rock had found its way to the museum in Jermyn Street.

[54] From Stuðill, anything that steadies, a stud, prop, stay. A specific usage makes Stuðlar signify pentagonal basalt columns, and Stuðla-berg is a basaltic dyke (Cleasby). It is popularly opposed to Mó-berg, “a kind of tufa,” properly Palagonite, from Mór, a moor or peat-fuel.