SOSMAN AND MERWIN : DATA ON PALISADE DIABASE 395 



could have been subjected, and the minmium temperature at which 

 the diabase could have flowed. At the time of intinision, there- 

 fore, either the fusion temperature of the arkose minerals was 

 greatly raised by the conditions then existing, or the fusion tem- 

 perature of the diabase was lowered, or both causes acted together. 



From experiments not yet published, we know that there is a 

 considerable net volume increase accompanying the fusion of the 

 diabase; its mean temperature of fusion would therefore proba- 

 bly be raised by increased pressure. About the volume change 

 of the arkose minerals we know nothing ; but from geological evi- 

 dence the depth of overlying rock was probably not great, in 

 comparison with the depths and pressures necessary to produce 

 large changes in their fusion temperatures. 



The most likely agent that suggests itself to bridge the gap 

 between the laboratory fusion temperature of the arkose and 

 that of the diabase is the water which the evidence of the sur- 

 rounding rocks shows to have been present in the magma. Its 

 effect in lowering the fusion temperatures of silicates is quantita- 

 tively unknown, but qualitatively we know it may be very large. 

 Other factors of perhaps less importance are the boric acid now 

 found in the datolite of surrounding rocks, and the carbon dioxide 

 and fluorine which may have been present. 



The foregoing investigation may serve to emphasize a point 

 which is all too frequently overlooked in current speculation on 

 the fusion of rocks on the basis of the laboratory data now avail- 

 able. The conditions of fusion, differentiation, intrusion, and 

 crystallization may have been modified, not slightly but pro- 

 foundly, by volatile colnponents of which only occasional traces 

 are retained in the rock as we find it at the surface of the earth. 

 To draw final conclusions, then, on the basis of any property 

 possessed by a particular rock type or even by a particular speci- 

 men may lead us widely astray. This is not to say that the prob- 

 lems involving the effect of volatile components can not be 

 solved, for experimental methods have been and are being worked 

 out to handle them; but until we know in what directions and 

 to what extent these volatile ingredients ha^'e modified the con- 

 ditions of rock formation, extended generalisation is hazardous 

 and of doubtful utility. 



