1890. ] Recent Literature. 555 
that, on the whole, the ‘‘ Cours de Mineralogie’’ is better adapted to 
the wants of the well-rounded mineralogist than any other single book 
published. A table showing the relations between the crystallographic 
symbols of Lévy, Miller, Naumann and Dana, and an index of thirty- 
five pages, complete the volume. Before concluding tĦħis notice, it 
should be mentioned that the author finds no existing method of min- 
eral classification satisfactory to himself. He divides the minerals into 
four groups, as follows: 1, the elements of rocks; 2, the elements of 
mineral deposits ; 3, metallic minerals, and 4, combustible minerals ; 
and uses this classification as the basis of the systemazc portion of his 
book. 
Lévy’s Structures et Classification des Roches Erup- 
tives.—Lévy’s small volume? on the classification and structure of 
rocks is so entirely argumentative that no satisfactory analysis of it 
can be given in these pages. It is directed against Professor Rosen- 
busch’s classification. Many instances are cited to show that the 
principles of this latter classification, when pushed to their legitimate 
consequences, must lead to the grouping together of rocks that have 
little similarity to each other, while, on the other hand, many that are 
evidently closely connected genetically must be widely separated in 
different groups. Lévy calls for a purely petrogr o classification 
of rocks, independent of geological consideratio e author’s 
cause would have appeared much stronger had his sate been less 
sprinkled with claims to priority over Rosenbusch in the proposal of 
terms descriptive of rock structure. The book merits close study as 
an appeal to petrographers to cut loose from theoretical considerations, 
and to make their classifications, for the present at least, expressions of 
observed facts. 
Thomas’s Ohio Mounds.’—In these two papers Dr. Thomas 
continues to maintain his thesis, already noted in our pages, that 
the earthworks of Ohio were built by the ancestors of the Red Indians 
of historic time. In the second of the two papers, the Cherokees are 
shown to have been mound-builders since the advent of the whites; 
and our author tries to trace them northward, connecting them with 
the monuments in West Virginia (near Charleston), and also with the 
traditional Tallegwi. To this end the Walum olam is invoked to show 
2M. Levy; Structures et Classification des Roches Eruptives. 95 pp. Paris, 1889. 
3 Thomas, Cyrus: The Circular, Square and Octagonal Earthworks, of Ohio. Pp. 
iii. The Problems of the Ohio Mounds. Pp. 33-+ii. Washington Bureau of 
Ethnology, 1889- 
