590 CARAVAN Ss 
designated as a type which has certain characteristics. _However, 
the only method of nomenclature which is logical is that which 
takes into account the fact of gradation and stages between all 
rock varieties ; that types exist only as descriptions or specimens 
selected by man, not by nature; but abundance is determined 
by nature. All this 1s recognized in sthe scheme) given aw) sline 
definite forms which are selected for names are those which are 
abundant. This applies as well to the original forms of rocks 
as to their altered varieties. Had the varieties of rocks inter- 
mediate between those abundant kinds which have been given 
names been the abundant ones, rather than those to which names 
have been assigned, these would have been the rocks which 
should have been given names, and which in all probability 
would have been given names during the first period of the devel- 
opment of the science of petrography. 
How shall demands for exactness be met?—But the question 
now arises as to how the demands of the petrographers for 
exactness shall be met without introducing a new independent 
name the moment a slightly different variety of rock is discov- 
ered, however small its mass. This demand may be met by the 
general application of special usages below given. The majority 
of the abundant kinds of rocks were early assigned names. To 
less abundant kinds of rocks intermediate between the more 
abundant kinds, names compounded from the simple names may 
be used; for example, granodiorite, trachydolerite, trachy- 
andesite." To either the simple or the compound names may be 
prefixed mineralogical qualifiers, thus further compounding them ; 
for example, quartz-diorite, olivine-gabbro, analcite- basalt, 
augite-trachydolerite. With these simple or compound names, 
geographical qualifiers may be used. If the rock is so abundant 
and definite as to require a specific name, the geographical 
qualifier may be compounded with the more general name, as 
Hellefors-diabase. If, however, the idea is exactly to designate 
the rock occurring at a particular locality, without implying that 
‘Italian petrographical sketches, V, by H S. WASHINGTON: JouR. GEOL., Vol. 
V, 1897, pp. 365, 366. 
