THE ANALCITE BASALTS OF SARDINIA 749 



lower than that of any of the three rocks analyzed by me. His 

 Fe203 is also undoubtedly too high (apart from the non-determina- 

 tion of FeO), presumably because zinc was used for the reduction 

 preparatory to titration with permanganate, part of the TiOz being 

 thus reduced and estimated as Yejd^. His alkalies are also low, 

 and it seems possible that they have been interchanged. This 

 analysis may therefore be left entirely out of consideration. Dan- 

 nenberg {op. cit., p. 50) gives the silica percentage of the Scano rock 

 as 44.16, which agrees very well with my results, especially as a 

 little siKca would be recovered from the alumina precipitate. 



Occurrence of analcite. — The results of these analyses render it 

 highly improbable that the leucitic mineral is really leucite, and 

 indicate clearly that it is analcite. To test this, the alkahes were 

 determined in the portion of the Scano rock soluble in warm dilute 

 hydrochloric acid. This gave Na20= 2.66 and K20=o. 12 in per- 

 centages of the rock; that is, more than half the soda and about 

 one-sixth of the potash. Some of this soda is probably derived 

 from the glass base, but its amount is so great, and that of potash so 

 small, that it leaves no doubt that the rounded isotropic crystals 

 are analcite. The entire absence of twinning lamellae and optical 

 anomalies is also in harmony with this supposition, as it is well 

 known that analcite, especially when not freely crystallized in 

 cavities, is much less prone to show such phenomena than is 

 leucite.^ 



The very fresh condition of the rocks analyzed precludes the 

 idea that the analcite is secondary and replaces an original leucite — ■ 

 a supposition which is often invoked apparently only because of a 

 disbelief that a hydrated, zeolitic mineral like analcite can be pri- 

 mary. The occurrence of primary analcite has been shown by 

 several petrographers,^ and there is also considerable experimental 

 evidence, which cannot be gone into here, that confirms the view 

 that such a primary character is not only quite possible, but very 



' The absence of cleavage does not militate against the idea here advanced, as it 

 is not a strongly marked characteristic of small analcites in thin section. 



^ W. Lindgren, Proc. Cal. Acad. Sci., Ill (1890), 51; L. V. Pirsson, Jour. GeoL, 

 IV (1896), 686, and U.S.G.S., Bull. 237 (1905), 154; W. Cross, Jour. GeoL, V (1897), 

 684; J. W. Evans, Q.J.G.S., LVII (1901), 38; G. W. Card, Rec. G.S.N.S.W., VII 

 (1902), 100; A. Lacroix, Mater. Min. Mad., I (1902), 197. 



