178 Williams — Petrography of Fernando de Noronha. 



gases around the flame radiate a considerable amount of heat 

 to the pile, and yet give off no light. Dr. Thomson also found 

 that only about 15 per cent of the heat of combustion was ra- 

 diated, the remaining 85 per cent being lost by convection. 

 The true efficiency of an oil lamp as a light-making machine 

 is therefore only about 0*3 per cent, or one-tenth that of an 

 incandescent lamp. 



It will thus be seen, as has already been pointed out by sev- 

 eral writers, that the incandescent lamp, though an immense 

 improvement on gas and oil lamps, is still far from being an 

 efficient light-making machine. Whether a light machine 

 will ever be found whose efficiency is much greater, it is of 

 course impossible to say, but it is a question which merits the 

 serious attention of investigators. 



Art. XX. — Geology of Fernando de Noronha. Part II. 

 Petrography ; by G-eobge H. Williams.* 



What has heretofore been published in regard to the petrog- 

 raphy of the island of Fernando de Noronha is inconsiderable, 

 but this fact is due to the difficulty in obtaining access to 

 the locality and the consequent rarity of material rather than 

 to any lack of interest in the rocks themselves. 



The volcanic nature of the island and the true character of 

 its prevailing rock (phonolite) were recognized by Darwin in 

 1832. f The place was visited by the Challenger party in 1873, 

 from one member of which (Willemoes-Suhm) Giimbel ob- 

 tained a single specimen of the phonolite which he analyzed 

 and examined microscopically in 18794 Other specimens col- 

 lected by Mr. J. Y. Buchannan, also a naturalist of the Chal- 

 lenger expedition in 1873, from islands of the Fernando group, 

 although not from Fernando de Noronha itself were described 

 by Abbe Renard in 1882.§ He figures and gives the micro- 

 scopic characters of a phonolite from St. Michael's Mount (Sella 

 a Grineta)| and of a nepheline basalt from Rat Island (Ilha 

 Rapta) with an analysis of the latter. He also mentions a feld- 

 spar basalt from Platform Island (Sao Jose), and a calcareous 



* For Part I, see page 145 of this volume. 



f Geological Observations on Volcanic Islands, London, 1844, p. 23. 



% Tschermak's Mineralogische und Petrographische Mittheiluogen, ii, p. 188. 

 1879. 



§ Notice sur les roches de l'ile de Fernando Noronha, Bull. d. l'Acad. roy. de 

 Belg. (3) III, No. 4. 1882. 



|| It is probable that the phonolite specimen studied by Giimbel came from this 

 locality, as the Challenger party appear not to have landed for scientific purposes 

 on Fernando de Noronha itself. (Cf. Wyville Thompson's narrative, The Atlantic,, 

 II, p. 115. 1877.) 



