H. Stanley Jevons — Nomenclature of Igneous Rocks. 311 



group possessing well-recognised characteristics. It is inevitable that 

 the meaning of a name should be altered from time to time — some- 

 times narrowed, though generally extended ; but every such alteration 

 must be gradual, taking place step by step. The reason for this lies 

 in human nature. A radical alteration in the meaning of a name 

 involves such a revolution of ideas and habits that most men refuse 

 to accept any such proposal ; and its adoption by only a few simply 

 leads to endless confusion. On the other hand, a slight alteration of 

 meaning requires but a trifling readjustment of ideas and habits; 

 and, therefore, so long as the change is clearly justified, it is soon 

 recognised as convenient and gradually comes into use. 



Whilst these objections clearly prevent the adoption of such 

 names as lujaurite, litchfieldite, etc., to replace nepheline-syenite, 

 there is one name, foyaite, which would appear at first sight to 

 escape them. It has been suggested, and sometimes used, as 

 a synonym for nepheline-syenite, and few authors, if any, now 

 restrict its meaning solely to the type of rock originally described 

 by Blum under that name. A little investigation shows, however, 

 that even this name has not had an application wide enough to 

 allow of its extension over the whole group which I propose to call 

 midalkalites. Rosenbusch uses it to include only what might be 

 called the normal types of nepheline-syenite ; those which are poor 

 in coloured constituents, and in which the alkali felspars are both 

 abundant, the potash felspar predominating over the soda felspar, 

 rather than the reverse. This use of it excludes such types as 

 Litchfieldite, Laurdalite, Lujaurite, and many others, but includes 

 Brogger's and Derby's Foyaites. At the same time it appears 

 probable that the majority, if not all, of those authors who have 

 used foyaite as synonymous with nepheline-syenite, have had in 

 their minds only those normal types included by Rosenbusch under 

 that name. The name is therefore applied to a definite group of 

 rocks closely allied to one another, and it would be a contravention 

 of the principle above stated to apply it to such widely difi"ering 

 rocks as laurdalite, litchfieldite, borolanite, leiicite-syenite, and sodalite- 

 syenite. It will, indeed, be found convenient to retain the name 

 foyaite with Rosenbusch's meaning as marking a subfamily. 



The only course left me was to coin a new name. A descriptive 

 name seemed likely to be most convenient, and I therefore chose 

 midalJcalite. The fact that all other family - names are of the 

 non-descriptive kind, and that I do not at the same time propose 

 descriptive names to replace them, is no valid objection to the 

 introduction of one such name. If this one prove a success, others 

 can be added from time to time as desired. 



Two other family-names require a brief notice. Desiring to 

 reduce the number of existing family-names as far as possible, 

 I have followed Rosenbusch and grouped together the soda and 

 potash rocks, so that the theralites and shonkinites form one family, 

 and the ijolites and missourites another. Instead of retaining the 

 double name for each family, however, as he does, I have selected 

 one of the names in each case to cover both the soda and potash 



