LINOSA AND ITS ROCKS 



29 



includes the very small fragments of basalt which seem to be always 

 present. They were made on air-dried material, the specimens having 

 been kept in a drawer for over a year prior to the analysis. None of 

 the minor constituents, except titanium dioxide, were determined, 

 but both manganese and nickel were present, and apparently in 

 amounts comparable to those shown by the preceding analyses. 



II 



la 



II a 



SiOz. . 

 AI2O3. 



FeO... 

 MgO.. 

 CaO.. 

 Na^O. 

 K2O.. 

 H,0 + 

 H2O- 

 CO2.., 

 TiO^.. 



47-43 

 17 .20 

 4.2- 

 5-27 

 4.85 

 7-56 

 3-53 



2.42 

 3.12 

 none 

 3.00 



39.00 



15-58 



6.13 



3-1^ 

 6 55 

 6.82 



3 -22 



0-59 

 6.03 

 8.18 

 1.83 

 2-59 



50.16 

 18.19 



4-44 

 5-57 

 5-13 

 8.00 



3-73 

 1 .60 



3-17 



46.66 

 18.64 

 7-33 

 3-72 

 7-83 

 8.16 



3-85 



3.10 



100.09 



99-63 



99-99 



I. Gray tuff. Monte Levante, Linosa. 



II. Yellow tuff. Monte Pozzolana, Linosa. 

 la. Analysis I calculated free from water. 



Ila. Analysis II calculated free from water and carbon dioxide. 



These analyses present some features of interest. The silica of I 

 is about the same as that of the lavas, while that of II is very much 

 lower. Also the alumina of I is distinctly higher than in the lavas, ^ 

 while that of II is about the same as in these. In view of the evi- 

 dently very considerable weathering which these tuffs have undergone, 

 the much lower ratio of ferrous to ferric oxide, as compared with the 

 figures shown by the fresh lavas, is not surprising. Magnesia is very 

 decidedly lower in I than in the majority of the lavas, while in II it is 

 but slightly so, but the figures for lime as well as for the alkalies and 

 titanium dioxide do not differ materially from those in the fresh rocks. 

 The presence of considerable water only driven off at a high tempera- 

 ture, and of carbon dioxide in II, is readily understood, but the high 



I This high alumina, taken in connection with the low magnesia, might be 

 thought to be due to the common analytical error of partial precipitation of the mag- 

 nesia with the alumina, owing to the presence of insufficient ammonium chloride- 

 But this was specially guarded against, and the absence of magnesia from the 

 precipitate by ammonium hydroxide was definitely proven. 



