A. Wade—Composition of Igneous Rocks. ivt 
the name antarcticus is retained. To this variety belongs also the 
Bolivian species Sp. Chuquisaca, Ulrich. Some of the South African 
specimens figured by Sharpe (14, pl. xxvi, figs. 1, 3, excl. cet.) under 
the names Sp. antarcticus and Sp. Orbignyt are referred to the typical 
form of Sp. arrectus. The species Sp. Hawkinsi (Morr. & Sh.) from 
the Falkland Islands is chosen to designate another variety of 
Sp. arrectus, to which also Sp. Vogeli, V. Ammon, from Brazil, is 
assigned. According to Scupin, therefore, the South African, 
Brazilian, Bolivian, Falkland Islands, and, we may add, Argentine 
forms known under the diverse names of Sp. antarcticus, Orbigny?, 
Chuquisaca, and Vogeli may all be included in the one North 
American species Sp. arrectus, which is closely allied to the European 
Sp. primevus. 
In the Devonian beds of Tassili in the Sahara, Haug (11) records 
Spirifer cf. Rousseaui, M. Rouault, which, as he states, belongs to the 
same group. 
The occurrence of Zropidoleptus carinatus in the Bokkeveld Beis, 
though mentioned by Ulrich (9), is somewhat doubtful, no other 
paleontologist having observed it. 
(To be continued in our next number.) 
ViI.—Taxr Curuicat Composition or Ienrous Rocks: A New Mernop 
OF OBLAINING IT. 
By ArtHur Wapg, F.G.S. 
NY science which cannot express its conclusions by means of 
ys number must still be regarded as in its infancy. Such an 
expression is the only test of precision and accuracy of thought on 
the part of its workers, and such important characters only develop, 
as in the human mind, when the romance and rashness of youth have 
given place to the more cautious and matter-of-fact features of 
maturity. It is therefore a sign of healthy development when those 
engaged in the study of petrology begin to turn their attention towards 
careful numerical calculation as a basis of classification. 
This has been done in a very elaborate fashion by the authors of the 
new ‘‘ Quantitative Classification of the Igneous Rocks,” and whilst 
reserving our opinion with regard to the ultimate value of this 
new American scheme, there is no doubt that it emphasises the 
importance of certain features which we in Britain are apt to hold too 
lightly. The most important of these is the absolutely essential nature 
of obtaining some fairly accurate idea of the chemical composition of 
a rock before proceeding to assign to it a position in any rock-group or 
family. ‘The making of a complete chemical analysis of a rock is 
a somewhat tedious and delicate process requiring much technical 
1 Cross, Iddings, Pirsson, & Washington: ‘‘ Quantitative Classification of the 
Igneous Rocks ’’ ; Chicago, 1908. 
