DESCRIPTION OF ROCKS AND ANALYSES 25 
companied by rising content of alumina, water, and partly lime, 
by lowering of the content of alkalies (soda) and silica and by 
lowering of the specific gravity of the rock. Only in this way the 
7 early cristallisation of thomsonite is intelligible and the forming of 
its pseudophenocrysts is drawn into true light. The real content of 
soda zeolites in the rock is very low, as showed by the analysis 
‚figures, yet amongst the cavity mineral they predominate, and the 
relative soda content in each of them exceeds the soda content in 
the albite aione. 
That something of this kind has worked out the actual stage of 
the rock, whose mineral composition is not due to the weathering 
process, finds a confirmation in the mode of olivine-alteration and 
perhaps in the lowered birefraction of the thomsonite mineral. Yet 
this process does not exclude an original rock of the calc-alkali 
series, if not the order of zeolite-separation might strengthen an 
alkali affinity. The olivine rich in ferrous silicate gives identical 
hints. The statement by Holmes,! that among the volatile fluxes 
of the alkali series carbon dioxide was a dominant member, cannot 
be strengthened in this case, no carbonate mineral having been 
found among the cavity fillings and only traces of them in the rock. 
In consideration of this the writer resolved to rank the rock among 
the trachydolerites of the alkali series unless a hasty examination 
of both the rock and the analysis should rank her with the common 
basalts. Except alumina and alkali content and the lower percen- 
tage of titanium dioxide, perhaps individual, the analysis agrees well 
with the analysis of the olivine-trachydolerite of Mt. Halvdan. 
b. Wilkitski-island. This little island east of the Jeannette and 
Henriette islands is represented by the material brought home by 
the Hydrographic Expedition of lieut. B. Wilkitski in 1913. There 
1 A. Holmes, On the tertiary volcanic rocks of Mozambique. Q. J. G. S. 72. 
1917. p. 253; this statement agrees quite well with the fact of increasing lime- 
content to the end of the postmagmatic stage, the present rock going not farther 
then to lime-silicate and not reaching the carbonate stage. 
