No. 411.] REVIEWS OF RECENT LITERATURE. 237 
with a view to determining the mode of origin of certain rock masses. 
He lays off on a system of rectangular coordinates the proportions of 
the chemical constituents present in the known members of a series 
of rocks derived by the differentiation of the same magma, and draws 
lines through the corresponding points in the figure. Then by inter- 
polation the composition of intermediate members of the series may 
be determined by inspection. When the lines passing through the 
points representing the proportions of the constituents present in 
the various rocks are straight lines, the series concerned is a “ linear 
series," that is, the rates of change in the constituents are constant 
and their proportion is determined by the percentage of SiO, present. 
This is a “rock series" in Brógger's sense. Such a rock series, 
however, Harker thinks nonexistent. The normal diagram of a 
*rock series" consists of curved lines, some of which are concave 
and others convex. Sometimes when a rock’s analysis is plotted the 
discovery is made that it does not fit into the diagram of the series. 
Such abnormal rocks may have been formed by the admixture of two 
magmas, or by the solution of foreign rock fragments in a normal 
magma. In either case the resulting mixture possesses a peculiar 
composition, the plotting of which does not fit into the diagram. 
For instance, the lavas of the Lassens Peak district, comprising 
basalts, andesites, dacites, and rhyolites, form a rock series the com- 
position of which may be represented by a well-defined diagram. The 
quartz-basalts, however, refuse to adapt themselves to the scheme. 
Their abnormal composition is clearly brought out by the plotting. 
In their content of lime and potash they do not differ notably from 
normal rocks of the same silica percentage, but they show a 
marked deficiency in alumina and ferric oxide and to a less degree 
in soda, . Magnesia and ferrous oxide are in excess. 
The plotting of analyses in the manner indicated by the author, 
and the interpretation of the diagrams thus made in the way out- 
lined above, form a ready means of detecting rocks of abnormal 
composition in a set of normal rocks from the same petrographic 
province. The diagrams thus may be made a means of aiding 
rock classification, since they enable one to exclude from dis- 
cussion those rock species which can have no part in any series of 
related rocks, and which if discussed together with normal rocks 
would give rise to difficulties hard to overcome in any scheme of 
classification. The author declares that a natural classification 
“must be based, confessedly or implicitly, upon fundamental genetic 
considerations, and primarily upon the mode of operation of the 
