PRESIDENTIAL ADDRESS. 379 



extreme in southern Arran, occur the analcime-dolerite sills of Clauchland and 

 JJippin. 



The regional basic dykes, which are mostly posterior in age to the sills, 

 exhibit more variety of composition. Some with abundant porphyritic felspars 

 resemble the Markle type of dolerite, and there are others of mugearitic nature, 

 but these are only a minority. In Argyllshire there are basic dykes with purple 

 pleochroic augite, and even some of camptonite and monchiquite; but these latter 

 at least I should exclude as being probably of late Palaeozoic age. 2 The un- 

 doubtedly Tertiary dykes, however, exhibit a variety which can be explained 

 only as the result of repeated differentiation. The distribution of some of the 

 groups indicates the existence at this late stage of subsidiary centres of dif- 

 ferentiation, distinct from the plutonic centres. Thus, trachyte dykes are 

 found especially throughout a tract extending from the south-western part of 

 Nkye through the middle of Argyllshire, while there is an isolated area of these 

 dykes about Drynoch, on the opposite side of the Skye mountains. Here we 

 have an evidently alkaline type. On the other hand, there are rocks which, 

 taken by themselves, must be assigned to the calcic division. Augite-andesites, 

 for example, are well-known, especially in parts of western Argyllshire, in Arran 

 and the Cumbraes, and in the outlying districts of the North of Ireland, Angle- 

 sey, and the north-east of England. That these rocks have arisen as products 

 of a subsidiary differentiation we have in some cases almost ocular demon- 

 stration ; for in Arran and elsewhere augite-andesites are found in remarkably 

 intimate association with complementary types, often pitchstones of alkaline 

 composition. 



Even from so brief and imperfect a sketch we may, I think, draw some 

 conclusions which have a wider application. This province exemplifies at once 

 the two main tectonic types, and also comprises representatives of the two great 

 branches of igneous rocks. Those rocks which are related to broad movements 

 of Atlantic type indicate a parent magma of decided, though not strongly 

 marked, alkaline nature; while those related to local movements of Pacific type 

 clearly come from a calcic magma. There are some facts which suggest that 

 the rocks tend to become more alkaline as we recede from the chief centres of 

 activity, and this suggestion applies to some calcic as well as alkaline groups of 

 rocks. Finally, it appears that the relative simplicity of arrangement was dis- 

 turbed at a late stage by the effects of subsidiary differentiation, the province 

 tending then to break up into districts related to new centres. Operating upon 

 an initial magma not very strongly characterised, this later differentiation has 

 even given rise to aberrant rock-types which overstep the petrographical boun- 

 dary-line between the two branches. 



Pedogenesis and Systematic Petrography. 



From such considerations as I have hastily passed in review, it is evident 

 that a survey of igneous rocks as they actually occur in the field leads to a con- 

 ception of their mutual relationships very different, from that embodied in the 

 current schemes of systematic petrography. It may be of some interest, in con- 

 clusion, to expand this remark a little further, although I am sensible that in so 

 doing I lay myself open to the charge of vain speculation. 



From the petrogenetic point of view, the most fundamental division among 

 igneous rocks is that between the alkaline and calcic branches. This result, 

 independently arrived at on petrographical grounds by several authorities, 

 seems to be firmly established by the broad distribution of the two branches 

 in different regions of the globe. But, if this argument be admitted, it follows 

 that the next step in a natural grouping of igneous rocks should be suggested 

 by a comparison of the characteristics of the various provinces into which the 

 great regions divide. Many of these provinces have now been partly studied, 

 and their special characteristics can often be expressed in concise terms : e.g., 

 among alkaline rocks the relative proportion of potash to soda may be a char- 

 acteristic common to a whole province. More precisely, by averaging the chemical 

 analyses of the chief rock-types, weighted according to their relative abundance, 

 it is possible to calculate approximately the composition of the parent-magma of 



2 A like remark applies to the highly alkaline dykes of the Orkneys, which 

 do nut agree even in direction with the Tertiary suite. 



