Augite of Haledkala, Hawaiian Islands. 121 



a mineral which must be regarded as but doubtfully 

 pyroxenic, and which does not appear in the recasting of 

 the best analyses of the non-aluminous pyroxenes. 



The molecular composition of the Haleakala augite, on 

 the basis assumed by us, and with titanium dioxide reck- 

 oned with silica, is as follows : 



CaMgSi,O c 69.12 



CaFeSi„0 6 15.13 



NaFeSi 9 6 5.08 



MgSiO s 1.90 



FeSi0 3 0.40 



(Al, Fe) 2 3 8.65 



100.28 



The Si0 2 + Ti0 2 demanded by the bases on this inter- 

 pretation is (molecularly) 0.828 or 49.68 per cent (reck- 

 oned as silica), while there was actually found 0.819 or 

 47.70 Si0 2 + 1.89 Ti0 2 , that is 49.59 per cent, a fairly sat- 

 isfactory agreement. The augite is thus essentially a 

 hedenbergite-diopside, with small amounts of acmite, 

 clinoenstatite, and alumina in solid solution. 



It may be instructive to give the molecular composition 

 as calculated on the basis of Zambonini's assumption, 

 that the A1 2 3 is present as a spinel (RO.R 2 3 ) in solid 

 solution. The composition on this basis is as follows : 



CaMgSi 9 6 58.97 



CaFeSioO e 12.65 



NaFeSi„0 6 • 5.08 



CaSiOs 6.61 



(Mg, Fe)0.(Al, Fe)oO s 12.23 



Si0 2 (excess) 4.20 



This interpretation is clearly not as satisfactory as the 

 preceding one, chiefly because of the presence of excess 

 silica, and partly because of the presence of the doubt- 

 fully pyroxenic wollastonite molecule instead of the 

 purely pyroxenic clinoenstatite molecule. The presence 

 of both of these somewhat exotic molecules is directly 

 brought about by the assumption of a spinel molecule 

 rather than free alumina and ferric oxide in solid solution. 



In calculating the mode from the norm in rock classi- 

 fication, it has been found to be very convenient to calcu- 

 late the norms of the mafic minerals present in the rock 

 (augite, hornblende, or biotite), the process being the 



