346 THE AMERICAN NATURALIST. [VoL. XXXVI. 
hydromica schists occupy the student’s attention when he might 
better be learning something more about gabbro than that it is a 
"coarse diabase.” The use of “syenitic” as an adjective synony- 
mous with hornblendic is no longer to be tolerated in an elementary 
text-book, in spite of “its wide popular use in this sense in this 
country.” This wrong usage of the word will never be done away 
with so long as it is taught in the text-books. 2: 
There are some inconsistencies in the classification, but they are 
not mischievous. There are also several statements concerning the 
composition and origin of a few rocks that will not bear close 
scrutiny. 
Since there « are no very elementary. text-books dealing with rocks 
published in English, the present volume has a place to fill. Itisa 
welcome aspirant for the honor of initiating beginners into rock 
study.. It would have been more welcome, however, had it.left 
some of its useless baggage behind. We hope Professor Tillman 
will revise the entire book, as there are many good points about it. 
In its scope it certainly “fills a want.” The volume contains 38 
crystallographic figures but otherwise is not illustrated. W.S. B. 
The Differentiation at Magnet Cove. — Washington! has. made 
new analyses. of some of the Magnet Cove plutonic rocks that lead to 
interesting conclusions. He finds that many of the earlier analyses 
are lacking in accuracy. The new analyses result in a change in the 
former classification of some of the rocks and seem to render neces- 
sary a few new rock names. From the analyses he calculates the 
composition of some of the important qe to be as in the table on 
opposite page. 
The covite was formerly irdd asa onsite. Its qualitative 
composition is close to that of theralite. The latter, however, con- 
tains a large quantity of nepheline, while. the former contains a 
comparatively . small quantity. On the other hand, shonkinite is 
melanocratic, while covite is mesocratic (that is, it contains about 
equal quantities of dark and light. constituents). . According to 
Rosenbusch's definition covite is a typical shonkinite, but Rosen- 
busch's shonkinite is an entirely different type from the prisia 
shonkinite of Pirsson. 
: The arkite. has been wrongfully designated a poni ti is most 
nearly allied to missourite, but it differs from this rock in the same 
way that covite differs from. shonkinite, że., it is a leucocratic 
; Comieco, cU Fourn, of Geols, vol ix (1901); p. 607.- 

