490 Prof. S. J. Shand — Saturation in Petrography. 



of the Absolute scale of temperature. If Mr. Scott's arguments were 

 extended to other things besides rock-classifications, nobody would 

 use a lamp, because it is far inferior to daylight, or a taxicab, 

 because it is not an aeroplane. 



Although I advocate separate chemical and mineralogical classifica- 

 tions for rocks, it does not follow that the mineralogical classification 

 must be devoid of chemical significance. It must, indeed, divest 

 itself of all attempts to express the percentage composition of rocks 

 in terms of oxides or arbitrarily chosen molecules, but in the act of 

 stating the proportions of various minerals in a rock we unavoidably 

 make some sort of a statement of its composition. Furthermore, the 

 investigation of the proportions which the constituent minerals bear 

 to one another involves the recognition of eutectic proportions, where 

 such exist. If eutectic ratios are to play a part in petrographic 

 system, it is in the mineralogical classification that they will find 

 expression, not in that which deals Avith ultimate chemical 

 composition. And if there be any other chemical principles in 

 operation, which express themselves in the presence or absence or 

 mutual relations of specific minerals, such principles must be taken 

 into consideration in the construction of the mineralogical groups, 

 since they enable us to make those general propositions which are 

 the soul of a classification. 



If the difference between the saturated and the undersaturated 

 rocks bears testimony to the operation of a general principle, namely, 

 that every magma tends to satui'ate itself with silica, then this 

 difference must find expression in our system. I believe in the reality 

 of the principle just enunciated, basing my belief primarily on the 

 evidence of the rocks themselves. I hold, therefore, that the 

 operation of this principle should govern the mineralogical classifica- 

 tion of rocks, and should be apparent in petrographical nomenclature. 

 The degree of saturation is easily observed, except in the case of 

 vitreous rocks, where all mineralogical criteria are equally inapplicable. 

 It is the " sure mark of many other properties ", such as the presence 

 or absence of specific minerals, the capacity of the magma to 

 assimilate foreign rocks, and the nature of the reactions thereby 

 induced; furthermore, it throws light on the ' cooling-historv ' of 

 the magma (which, as I read the facts, generally includes 

 a progressive increase, but sometimes a local decrease, in the degree 

 of saturation), and on the relations existing between the partial 

 magmas or differentiates of an igneous complex. For all these 

 reasons the criterion of saturation appears to me to be one wliich is 

 eminently adapted to be used in the construction of a mineralogical 

 classification of rocks. 



I come now to the more specific criticisms advanced by Mr. Scott. 

 My right to classify olivine as an unsaturated mineral is first 

 questioned. The obvious way to defeat my contention would have 

 been to adduce instances in which quartz and olivine are associated 

 in igneous rocks. Mr. Scott mentions one or two cases of this 

 association, and then plays into my hands by admitting that in one 

 case the olivine, in another the quartz, is "probably xenocrystic". 

 I am pleased indeed to have the additional evidence of Mr. Scott's 



