I 92 I. No. 8. ox THE ECLOGITES OF NORWAY. 7 



even the heavier rockmaking minerals, up to a specific gravity oi' 4.20, can 

 be separated without any difHculty. I prepared this solution simply by 

 neutralizing" thallium carbonate with equivalent amounts of formic acid 

 (85 pt) and malonic acid (concentrated aqueous solution) and by evaporating 

 the clear solution obtained until a piece of almandite (sp. g. 4.0) floated. 



For the separation the rocks were crushed up in a diamond mortar 

 and sifted, usuallv to a maximum diameter of o.i mm. The finest powder 

 was now removed by washing with water, and after drying the first 

 separation was made in a Brogger funnel. The specific gravity of the 

 liquid was determined, when the mineral in question Just floated, and once 

 more, when it just sank. These two values were not allowed to differ 

 bv more than 0.05. The average of them was taken as the specific 

 gravity of the mineral. A second separation was always necessary- before 

 the mineral proved pure by examination under the microscope, and some- 

 times the grains must still be crushed up to effect separation of double 

 grains or inclusions. 



The separation of the minerals in most cases was combined with a 

 mechanical analysis, or determination of the mode. Against this method 

 could be raised the objection that the result may be afîected by the removal 

 of the finest powder by washing', as softer minerals will be pulverized 

 more than the hard ones. I have not investigated how great errors may 

 possibly result here, but probably they are not very grave. There are 

 also other sources of error in this method, as the incomplete separation in 

 consequence of double grains, mostly ine\-itable even in the finest grain 

 practicable. 



Where analyses of pure constituents are available, the analysis of the 

 rock may often be omitted, the bulk composition of the rock then being 

 calculated on the basis of the mode and the mineral analyses. This method 

 was often practised in the course of the present work. The objection that 

 the results may be unreliable owing to errors in the mechanical analyses 

 is here of little importance, as in the case of the eclogites like so many 

 other banded or schlieric crystalline rocks the relative amounts of the 

 minerals itself are greatly variable. The figures obtained represent at all 

 events the composition of a mixture of the constituent minerals, which in 

 some parts of the rock must be present exactly in the proportions repre- 

 sented. The result may, moreover, be checked by calculating the specific 

 gravitv of the rock from those of its constituents in the mode, and com- 

 paring the result with the specific gravity of the rock directly determined. 

 Such calculations have always shown \ery close agreement. 



The determinations of the specific gravity of rocks were made in 

 pieces of 100 — 500 g bv weighing in air and water. Before weighing in 

 air the pieces were dried during the night in vacuo and the air bubbles 



' In one case the amount of the powder removed was determined to be 13 per cent. 



