NA TURE 



THURSDAY, SEPTEMBER i6, 1909. 



PRINCIPLES OF IGNEOUS PETROLOGY. 

 The Natural History of Igneous Rocks. By A. 

 Harker, F.R.S. Pp. xvi + 384. (London : Methuen, 

 1909.) Price I2S. 6d. net. 



FOR some time the need of a text-book of general 

 petrology in English has been acutely felt, owing 

 to the rapid progress which the science has made in 

 theoretical subjects and the inaccessibility of many 

 of the original memoirs to students and teachers. 

 Most text-books treat the subject from a purely 

 descriptive point of view, and the speculative develop- 

 ments are kept in the background. In fact the litera- 

 ture of descriptive petrology is now so large that an 

 attempt .to extract the general conclusions to be 

 deduced from the observations becomes of greater 

 importance than merely to add to the number of 

 ascertained facts. In a book of about four hundred 

 pages Mr. Harker endeavours to meet this demand, 

 and has covered so wide a range and compressed so 

 much information into this brief space that he has 

 achieved a very large measure of success. 



The work is confined to the investigation of the 

 principles of igneous petrology, and the metamorphic 

 and sedimentary rocks are excluded from its scope. 

 The title fairly expresses the aims of the book ; it is 

 an attempt to treat of igneous rocks in the manner 

 of natural history, taking account not only of their 

 minute anatomy and structure, but also of their dis- 

 tribution, their range in time, and their genetic asso- 

 ciations. Of late years there has been a plethora of 

 unnatural systems of petrology, based on purely 

 arbitrary lines. On the best known of these (the 

 " quantitative classification ") the author passes severe 

 judgment, and returns to the broader treatment 

 followed by Rosenbusch, Brogger, Judd, and Teall. 

 That combination of geological reasoning with petro- 

 logical analysis which has always characterised the 

 English school of petrologists finds able expression in 

 the pages of this treatise. 



The geographical distribution of recent volcanic 

 rocks is considered in the introductory chapters, and 

 the remarkable association of " Atlantic " and 

 " Pacific " rock types with certain classes of tectonic 

 developments is made the foundation of an appeal for 

 a natural classification. Undoubtedly the facts are 

 most impressive, though we are entirely in the dark 

 regarding their causes ; and they afford the clearest 

 indication that in time we shall be in possession of 

 natural systems of petrography. We learn incidentally 

 that the first sketch of this grouping was presented 

 by Mr. Harker, though it has become more generally 

 known through the treatment of Dr. Prior and Prof. 

 Becke. The Tertiary volcanic rocks of the Inner 

 Hebrides are placed among the Pacific types, a con- 

 clusion by no means easy to accept, and one which 

 may need to be revised at some future date. 



The application of physical chemistry to the crystal- 

 lisation of magmas is taken up in considerable detail. 

 This part of the book is practically a summary of 

 Prof. Vogt's papers, and leaves nothing to be desired 

 NO. 2081, VOL. 81] 



in lucidity of exposition and accuracy of statement. 

 For students it will be of the greatest value, as the 

 original memoirs are loaded with detail, and much too 

 voluminous for their use, and no satisfactory account 

 of them has hitherto appeared in English. We must 

 confess, however, to a certain feeling of disappoint- 

 ment in reading them, a sense of incompleteness. 

 The a priori principles are laid down in a very satis- 

 factory fashion, but the applicability of these prin- 

 ciples to the actual concrete facts of rock structure 

 and history, which we had a right to expect from a 

 geologist of Mr. Harker"s wide experience, is dis- 

 missed with scant treatment. The difficulty of inter- 

 preting the history of crystallisation in the commonest 

 rocks in terms of the theoretical principles laid down 

 is sure to face the student at an early stage in his 

 studies. We miss in particular any reference to the 

 work of Schreinemaker, who has proved by analytical 

 investigations that in a solution of three minerals in 

 one another, of which two can form mixed crystals, 

 while the third is independent, there are many possible 

 schemes of crystallisation ; one mineral may separate 

 out completely at an early stage, and the crystallisar 

 tion of any substance may be interrupted or repeated. 

 If there are more components, or if we allow for the 

 influence of dissolved gases in the magma, the problem 

 becomes much more complicated; but it is a rehef to 

 find that as the theory is better understood the dis- 

 crepancies between it and the observed facts seem to 

 diminish. 



The subject of " hybridism," or mixed rocks, is 

 discussed in a brief chapter, which sums up in admir- 

 able fashion the results of the author's work in this 

 difficult field. He takes a middle position between 

 the schools of petrology which deny that igneous 

 rocks dissolve sediments or older rocks with which 

 they come into contact, and those which hold that 

 such processes go on on a large scale, and that many 

 rocks generally regarded as of normal types are thus 

 produced. Mr. Harker "s field work enables him to 

 speak on this subject with great authority, and his 

 conclusions are so moderate and so firmly based on 

 sound evidence that he carries us with him in all 

 that he says. There is also a chapter on magmatic 

 differentiation, a subject on which it does not seem 

 possible to say anything that is new, and a very 

 interesting account of the mutual relations of asso- 

 ciated igneous rocks which, in our opinion, is the best 

 in the volume. The curves drawn on a very simple 

 graphic system show the variation in the components 

 of allied rock types, and are convincing that some 

 general principles must underlie the facts, though as 

 yet we have been unable to grasp them. The final 

 chapter on classification is unexpectedly brief, and 

 contains an admission that existing systems are 

 merely temporary stop-gaps, and a satisfactory classi- 

 fication must traverse the lines of all current group- 

 ings, and will require an entirely new nomenclature. 

 To us this appears unduly pessimistic, and we believe 

 rather that in petrology as in other sciences the 

 future will be the child of the past, and that real 

 progress will not involve the demolition of the older 

 systems, but will include them while giving them a 



N 



