Reviews — A. Harlm-'s Petrology. 515 



at once into his subject and teaches us just exactly what we want 

 to know about the shape, appearance, and general optical characters 

 to be observed in the microscopic study of minerals, and of the 

 rocks which they constitute. Then with a few passing words on 

 classification he begins his description of Plutonic rocks with the 

 Granites. 



He divides the igneous rocks into three broad classes, of which 

 the plutonic class needs no explanation; the intrusive rocks are 

 defined as "certain families of rocks" which "are met with almost 

 exclusively in the form of dykes, sills, laccolites of small dimen- 

 sions, and ' pipes ' of old volcanoes " ; and the volcanic rocks are 

 those which " have consolidated from fusion under superficial 

 conditions, i.e. by comparatively rapid cooling under low pressure." 



An uniform mode of description is adopted throughout. After 

 a few preliminary remarks on each great rock-group we are given a 

 list of the constituent minerals and their characters in the rocks 

 in question, an account of the structure of the rocks, of their 

 leading types, and, where necessary, of their special modifications. 

 The leading types, whenever possible, are defined from British 

 examples, but where the British Isles fail to supply the desired 

 information the whole world is laid under contribution. The 

 especial value of this part, and indeed of the entire work, is that 

 the description is written in almost all cases direct from the actual 

 specimens, and the author tells us with reserve of particulars and 

 variations which have not come within the wide range of his own 

 personal observation. Notwithstanding this, very full references 

 are given throughout, and this will make the book especially 

 valuable to the advanced student, as it gives him a mine of biblio- 

 graphic detail. 



A slight slip occurs under the Diorites, where the rocks of 

 Nuneaton are stated to be probably of Carboniferous age. As a 

 matter of fact these diorites, like the Cambrian sediments into 

 which they are intruded, are unconformably overlain by Car- 

 boniferous rocks, and are almost certainly not later than Ordovician 

 in point of date. Where similar rocks occur, as at Inchnadamff, 

 the Wrekin, and the Longmynd, they penetrate Cambrian or older 

 rocks only, and have never been found to touch Ordovician strata. 



The term Gabbro is not used quite in the sense suggested by 

 Teall, but appears to be confined to rocks possessing the diallagic 

 modification of augite. Two main types of ultrabasic rocks are 

 recognized, the picrites and the peridotites, the former "having 

 usually subordinate plagioclase," the latter being non-felspathic. 

 In dealing with the basic intrusive rocks Mr. Harker departs from 

 the usual, and, as we venture to think, the best, practice of British 

 petrologists in declining to use the term diabase for the altered 

 representatives of the dolerite family, and in applying it to ophitic 

 and granulitic rocks which "are holocrystalline and typically 

 non-porphyritic." Thus the word dolerite is reserved to designate 

 volcanic rocks of basaltic composition in which the distinction 

 between phenocrysts and ground-mass is not well marked. By 



