76 DB. H. S. WASHINGTON ON THE TITANIFEKOUS [Feb. I907, 



and the neighbouring submarine eruptions. It may be mentioned, 

 in this connexion, that two pantellerites of Pantelleria, with 69*79 

 and 66*07 per cent, of silica, yielded respectively 0*89 and 0-92 of 

 titanium-dioxide, which are unusually-high figures for such quaric 

 (acid) rocks. In a general way, titanium seems to follow the iron, 

 and runs inversely with the silica. 



Phosphoric pentoxide is distinctly abundant and, like titanium, 

 seems to follow the iron-oxides, the highest figures being found in 

 the rocks of Pantelleria and the neighbourhood. Chlorine (derived 

 from sea-water) is present in some of the Linosa rocks, but has not 

 yet been determined, whence is due in part the low summation of 

 No. 14. Nickel seems to be almost constantly present, even in the 

 Sardinian lavas, often in considerable amounts for this constituent; 

 and its presence is to be regarded as one of the characteristic, 

 though minor, chemical features of the region. Although only 

 occasionally determined as yet, its presence was established, as in 

 the Catalonian and Linosa basalts, by the faint greenish tinge of 

 the filtrate from the ammonia - precipitate, the amounts being 

 approximately those of the quantitatively -determined occurrences, 

 judging from the depth of colour. The small amounts of manganese 

 usually found do not call for remark. It is probable that copper, 

 the presence of which was sometimes indicated, exists only as traces, 

 although this is a point which is to be the object of future study. 

 Zirconium, barium, strontium, and sulphur, when looked for, were 

 found to be absent, or present in scarcely more than traces. 



Coming back to the abundance of titanium, the question naturally 

 arises as to what mineral or minerals carry it. It is clearly not 

 present in the felspars, and neither titanite nor ilmenite could be 

 detected in any of the thin sections. While the matter is still under 

 investigation, the available data indicate that a large proportion of 

 it is present in titaniferous magnetite, with probably somewhat less 

 in the augite, and little (if any) in the olivine. No separations have 

 yet been made of the rock-minerals, but an analysis of a dark-green 

 augite-nodule from a leucitic lava of Scano (Monte Ferru) gave 

 1*91 per cent, of titanium-dioxide, while an olivine from the same 

 locality contained only 0-10 per cent. 1 



As the augites of these rocks are almost, if not quite, colourless 

 in thin section, their richness in titanium may be doubted, as it is 

 commonly supposed that highly-titaniferous augites show a 

 brownish-violet colour. 2 It has, however, been shown recently 

 that the green augites of Central Montana, 3 and the pale-green or 

 colourless augites of Central Italy, 4 are markedly titaniferous, one 

 of the latter carrying 2*85 per cent, of titanium-dioxide. 



1 A black hornblende, found as loose crystals on Linosa, yielded as much as 

 8 - 47 per cent, of titanium-dioxide, the highest figure yet recorded for this 

 mineral. 



2 F. Zirkel, 'Lehrbuch der Petrographie ' vol. i (1893-94) p. 284; and H. 

 Rosenbusch, ' Mikroskopische Physiographie der Massigen Gresteine ' 4th ed. 

 vol. i (1905) 2nd half, p. 208. 



3 L. V. Pirsson, Am. Journ. Sci. ser. 4, vol. xx (1905) p. 39. 



* H. S. "Washington, ' The Roman Comagmatic Region ' Carnegie Inst. 

 Publ. No. 58 (1906) p. 334. 



