75 



becoming richer in iron and magnesia for these constituents do 

 not enter into the composition of felspars, and this, coupled with 

 the fall in temperature, led to the formation of augite. Finally, 

 tne interstitial matter consolidated. It occurs in two forms; as 

 a dark brown glass and also as an aggregate consisting of rods 

 of iron-ore, microlites of felspar, granules of augite, and a clear 

 substance which cannot be definitely identified. The brown glass 

 probably represents the interstitial matter undifferentiated. We 

 have no satisfactory analysis of this interstitial matter, but very 

 similar basalts occur in Franz-Joseph Land, and while preparing 

 a description of them in 1897 I isolated and analysed the palago- 

 nite which represents the interstitial matter of one of these rocks. 

 Neglecting the water, its composition expressed in minerals corres- 

 ponds very closely with a mixture of 40 per cent, of the soda-felspar 

 (albite), 22 per cent, of magnetite, 28 per cent, of olivine, 4 per 

 cent, of hypersthene, 6 per cent, of the lime-felspar (anorthite). 

 If a further differentiation had taken place soda-felsite and a rock 

 essentially composed of olivine and megnetite might have been 

 formed. 



" It follows from what has been stated that progressive 

 crystallisation in basaltic magmas produces different results under 

 different conditions of consolidation. What is the cause of this 

 difference? Some very important experiments on the consolida- 

 tion of artificial silicate-magmas have been carried out during 

 the past decade in the Geophysical Laboratory in Washington. 

 By modern appliances it is now possible to subject molten sili- 

 cate magmas or solutions — igneous rocks are merely frozen solu- 

 tions — to high temperatures, to measure the temperature at any 

 stage -as it falls, to maintain it at a definite point for any reason- 

 able length of time, and to examine the crystals present in the 

 mass at anv temperature by suddenly cooling the melt. It has 

 been proved that in artificial silicate magmas, more or less allied 

 to basalt, the ultimate result when the rock is entirely composed 

 of crystals, depends not only on the composition of the magma 

 but also on the rate of cooling; that with very slow cooling any 

 mixture of a certain pyroxene (diopside) and a plagioclase con- 

 taining albite and anorthite in equal proport ions (la bra dor t M 

 will consolidate as labradorite and pyroxene at 1200 Q C and 

 the felspar will be homogeous ; that with more rapid but not 

 too rapid cooling the felspars wi.ll be zoned with the central Darts 

 richer in anorthite than the margins, and that a mother-liquor 

 approaching albite in compo.^tion wi'Jl be formed ; and further 

 that with rapid cooling to 1200° C followed by crystallisation 

 at or just below that temperature the some resnl-* as that ob- 

 tained by verv slow cooling will be reached. It has also been 

 proved that olivine in certain mixtures nnd with certain rates of 

 cooling will be formed in the earlier stages and subsequently 

 recissolved. 



"These important experimental researches suggest that the 



