^ol. 57.] AXNIVERSAEY ADDRESS OP THE PRESIDENT. IxXXl 



The idea that petrographieal species have originated by differentia- 

 tion from homogeneous magmas, and possibly in the first instanco 

 from some one primordial magma, has been greatly developed during 

 the last decade of the century, especially by American and Nor- 

 wegian petrographers. Thus Prof. Iddings, in the introduction 

 to his important memoir on the Origin of Igneous Rocks, says : — 



' The object of the present paper is to give the writer's reasons for concluding 

 that all of the volcanic and other igneous rocks of any region are so intimate] y 

 connected together by mineralogical and chemical relations that they must 

 have originated from seme single magma whose composition may be different 

 in different regions ; and, further, that it is the chemical differentiation of this 

 primary magma which has given rise to the various kinds of igneous rocks.' 



The fact that the diverse igneous rocks of certain districts are 

 often bound together by common mineralogical and chemical 

 characters which distinguish them from the corresponding rocks of 

 certain other districts was clearly recognized by Prof. Judd in his 

 well-known paper on the "Volcano of Chemnitz, and subsequently 

 crystallized by him in the happy expression petrographieal 

 province, as applied to any district in which the igneous rocks 

 have certain common characteristics. The idea has been still further 

 extended and elaborated by Prof. Iddings, who sees in the common 

 characteristics the indications of a kind of blood-relationship or 

 consanguinity, which can only be explained on the assumption that 

 the different species of one and the same province have originated 

 by differentiation from a single homogeneous magma. 



Prof. Brogger, in his remarkable series of studies on the rocks 

 of the Christiania district, has still further generalized this idea, and 

 much of his work is directed towards the evolution of a genealogical 

 tree, in which the twigs shall correspond to the final products of 

 differentiation, the larger branches to some of the plutonic masses, 

 and the trunk to the primordial homogeneous magma. The idea is 

 a fascinating one : se non e vero, e ben trovato. But it must be 

 admitted that we know very little about the causes of the assumed 

 differentiation. These are supposed to be of two types: (1) those 

 which affect the liquid magmas, and (2) those connected with the 

 separation of the minerals. Magmatic differentiation is generally 

 regarded as the most important, but it is the type of which we know 

 least. Soret's principle, to which I have appealed, will, I fear, help 

 us very little, though it is undoubtedly a vera causa. Mr. Harker 

 has clearly shown that, as applied to a mass like the Carrock-Fell 

 gabbro, it breaks down hopelessly when subjected to a quantitative 



