R. A. Daly — Mechanics of Igneous Intrusion. 109 



Quartz 16-85 <fo 



Olivine 16-30 



Augite 14-18 



Hornblende 12-13 



Oligoclase 12'19 



Orthoclase __. __ 10-21 



Microcline 9-15 



Adular 7-96 



Sanidine _ 7-63 



Labradorite _. 6-28 



Barns has shown that, for silicates, the solidification contrac- 

 tion decreases in proportion as the thermal expansion decreases.^ 

 Since the thermal expansion is in direct proportion to the 

 decrease of density affecting a crystalline silicate passing to the 

 cold glassy state, it follows that the percentages of the table 

 are in nearly constant ratio to the percentage decreases of 

 density affecting the respective glasses when fused to a thor- 

 oughly molten condition. According to Barns, diabase loses 10 

 per cent in density in passing from rock at 20° C. to glass at 20° C. 

 and its glass loses 7*2 per cent in density in passing from 20° C. 

 to 1400° C, at which temperature it is very fluid. It is highly 

 probable, therefore, that olivine glass would lose in density 



something like 7*2 X — — = 11'7 per cent in assuming the 



same temperatm-e of 1400° C. The net result on that assump- 

 tion would be to give Fogo olivine, for example (spec, grav., 

 3-381) a specific gravity of about 2-50 at 1400° C. The spe- 

 cific gravity of gabbro at 20° C. becomes, when molten at 

 1400° C, changed to the following values :f 



Spec, grav, at 20° C. 



Spec. grav. at 1400' 



Holocrystalline gabbro. 



Molten glass. 



2-90 



2-42 



3-00 



2-51 



3-10 



2-59 



Assuming the approximate correctness of these figures, it is 

 possible to credit the flotation of molten globules of the oli- 

 vine substance in a highly fluid gabbro or basalt. The rock- 

 matrix would crystallize before the segregation or, at least, 

 would have attained an antecedent viscosity great enough to 

 support the more dense, because crystallizing, olivine. The 

 foregoing reasoning leads to the suggestion for actual experi- 

 mentation with olivine similar to that so successfully carried 

 out for diabase by Barns, Without such fusion tests it is not 



*Bull. U. S. Geol. Snrv., No. 103, p. 43, 1893. 

 t This Journal, April, 1903, p. 277. 



