Correspondence — G. W. Tyrrell. 335 



instead of directly towards the wind, either on account of its shape or 

 because the sand has been removed unevenly from below it. 



If a general expression be required for any wind-shaped stone, we 

 might speak of a ' ventifact ', on the analogy of artifact, sometimes 

 spelt ; artefact ', which is already in use for an object, such as 

 a palaeolith, fashioned by men, and of 'ventiduct', which has been 

 employed in architecture. John W. Evans. 



Imperial Institute. 



June 7, 1911. 



BEITISH PILLOW-LAVAS. 



Sir, — The brilliant paper on British pillow-lavas by Messrs. Dewey 

 and Flett in your May and June numbers is an illustration of what 

 may now be done in rational petrology by the collation and interpre- 

 tation of the great mass of facts accumulated on the descriptive side 

 of the science. The main points brought forward by them, i.e. the 

 existence of the spilitic suite, and the explanation of its association 

 with black shales, limestones, and radiolarian cherts, seem now to be 

 thoroughly established. 



One is obliged, however, to dissent from the view (pp. 242, 245) 

 that the spilitic suite is separate and distinct from, and, so to speak, 

 co-ordinate with the Atlantic and Pacific kindreds, as established by 

 Harker, Becke, and Prior. Using these terms merely as convenient 

 names for the two broad chemical divisions in igneous rocks, and 

 disregarding the much-disputed distributional assumptions on which 

 they rest, it seems to me that the Atlantic and Pacific branches cover 

 the entire field of igneous rocks. This is certainly the view taken by 

 Harker in his Natural History of Igneous Hocks (chap, iv), although 

 the terms Atlantic and Pacific themselves are based largely on the 

 distribution of Tertiary igneous rocks. The actual basis of the 

 classification, however, is chemical, and the above terms are due to 

 a probably too wide generalization as to the distribution of the groups. 

 If this is the case the spilitic suite is merely a subdivision of the 

 Atlantic branch, as its characters agree well enough with the 

 definition of the latter (Harker, op. cit., pp. 90, 91). 



Exceptions to that geographical distribution of igneous types 

 implied by the use of the terms Atlantic and Pacific are now 

 multiplying at such a rate that it may be necessary to drop those 

 terms in their petrographic sense. In that event future research may 

 show that the igneous rocks are divisible into more than two main 

 classes, distinguished by broad chemical and mineralogical characters, 

 and associated with various types of earth-movements. The spilitic 

 suite may form one of these classes ; but what I wish to point out 

 is that as at present defined it seems merely to form a part of the 

 Atlantic kindred, using that term in the sense that petrographers use 

 it, to indicate the great division of ' alkalic ' rocks. Messrs. Dewey 

 and Elett consider that a close parallel exists between the spilitic suite 

 and the analcite-bearing igneous rocks of the Scottish Carboniferous 

 (p. 209). If so, there seems no reason why the latter, or indeed any 

 well-marked group, should not be elevated to the rank assigned to the 

 spilitic suite by the authors. 



