H. Stanley Jevona — Nomenclature of Igneous Rockn. 305 



If objection were talceii to making the relative abundance of minerals 

 a factor in the nomenclature, I should answer that I believe it will 

 prove a useful step, as too little attention has hitherto been paid to 

 variations in the relative abundance of a rock's constituents, and 

 there exists a very simple method of determining this factor.^ 



Other properties besides mineral composition are occasionally 

 made use of as bases of subdivision, and further prefixes will be 

 required in naming the groups thus formed. When the subdivisions 

 are based upon structure, prefixes may be obtained by contracting 

 the names of the various structures. For example, pilotaxitic can 

 be contracted to pitaxo-, intersertal to inserto-, and so forth. It is 

 almost exclusively amongst the dyke-rocks and lavas that structure 

 is made the basis of subdivision, and names would be produced 

 such as mipegmo-rlujolite for a granophyric or micropegmatitic 

 rhyolite, or hypidio-hasalt for a basalt with hypidiomorphiu-granular 

 structure. 



Besides mineral and structure prefixes, others denoting some fact 

 in the chemical composition of the rock as a whole may be required. 

 Thus has- or hasi- would mean basic, ali- would signify alkaline, 

 and so on. These prefixes should generally come immediately before 

 the family-name, and should be separated from it by a hyphen 

 when they are used to name subdivisions of the principal families. 

 Thus the alkaline-syenite family, a group which may be ranked 

 as of equal importance with the normal or calc-alkaline syenites, 

 would be named alisyenite ; but the name of the basic subdivision 

 of the diorite family, if such a subdivision were made, would be 

 written basi-diorite with a hyphen. 



III. Index Minerals. 



The use of a family-name must evidently in every case imply 

 the presence in the rock to which it is applied of certain essential 

 minerals, which need not therefore be mentioned in the prefix. 

 These I term the index minerals of a family, because, being common 

 to all its members, they serve to point out to which family a new 

 specimen belongs. As an example, I may take the family-name 

 granite. Every rock so named, whether there be a prefix or not, 

 must contain the minerals quartz and orthoclase in abundance, and 

 these, therefore, are its index minerals. Granites usually, of course, 

 contain other minerals besides, and some or all of them may be 

 mentioned in the prefix. In the same way the index minerals of 

 gabbro are obviously plagioclase and monoclinic pyroxene. 



It is very important that accurate definitions of the index minerals 

 of all rock-families should be provided; but I am not in a position 

 to supply them at present. A few such definitions I have been 

 successful in framing, and I therefore feel confident that with 

 sufficient study every family can be properly defined. 



In the meantime, before this work is completed, the want of 

 accurate definitions will not seriously afi"ect the application of 



1 See A. Rosiwal: Yerhandlungen der k. k. geol. Reichs-Anstalt, 1898, p. 143. 



DECADE IV. VOL. VIII. — NO. VII. 20 



