102 



SCIENCE 



[N. S. Vol. XL. No. 1020 



ieal analysis of a rock which is to be classi- 

 fied or else an ability on the part of the petrog- 

 rapher to make a satisfactory comparison be- 

 tween the rock in hand and a similar type 

 which has been analyzed. The proper nse of 

 such a system will be beyond the powers of 

 many who now essay to name rocks. Prob- 

 ably some petrographers will object to Iddings's 

 propositions as impossible of application by 

 many who now endeavor to use the current 

 system. But increasing precision in the sys- 

 tem in any direction must have the same ef- 

 fect in some degree, and the development of 

 petrography can not long be held back for the 

 sake of a simplicity which can be maintained 

 only at the expense of accuracy and efficiency. 



Chemical data can be logically applied to 

 the development of the mineralogical system 

 only on the basis of our knowledge of the re- 

 lation between chemical and mineral com- 

 position afforded by the several thousand rocks 

 of which good analyses are now available. 

 This relation has been studied by means of the 

 significant molecular ratios between silica and 

 bases or between various bases or groups. 

 Osann's work is preeminent in this direction. 

 His ratios and triangular diagrams on which 

 they may be plotted express clearly certain 

 chemical characteristics of igneous rock 

 groups. But the application of such data to 

 the revision of systematic mineralogical classi- 

 fication is a very complex problem the solution 

 of which has not been attempted as yet. 



Another means of expressing relations be- 

 tween chemical and mineral composition is 

 the norm of the quantitative system, and Id- 

 dings's book will always be cited in the litera- 

 ture of systematic petrography as notable for 

 its well-thought-out and far-reaching plan to 

 reconstruct the qualitative or mineralogical 

 system on the same principles which underlie 

 the quantitative classification. If it is prac- 

 ticable to develop the mineralogical system on 

 these principles, this is a first, long step for- 

 ward in its evolution, to be followed by many 

 other improvements suggested by experience. 

 The reviewer's opinion as to the success of this 

 attempt is naturally subject to the charge of 

 prejudice, hence he contents himself in point- 



ing out the great importance of Iddings's 

 systematic propositions, whether they are 

 finally adopted or not. They deserve the most 

 careful, unbiased attention of petrographers 

 in any case. If it proves to be practicable to 

 develop the mineralogical system in the way 

 proposed, it will then be brought into desirable 

 harmonious relations with the quantitative 

 system. 



In Volume I. of " Igneous Eocks " Iddings 

 presented his modification of the " Qualitative 

 Mineralogical Classification " in the usual 

 tabular form. A quantitative element is made 

 prominent here by establishing five major divi- 

 sions based on the dominance of I., quartz ; II., 

 quartz and feldspar; III., feldspar; IV., feld- 

 spar and lenads ; V., lenads. The character of 

 the dominant feldspar and the abundance or 

 subordinate part of the ferromagnesian min- 

 erals gave subdivisions of the larger ones. 

 While the lines of this scheme were not made 

 precise, they served to divide many rock groups 

 or varieties of current usage. 



The underlying idea in this scheme is clear, 

 but the detail with which it is worked out in 

 Volume II. suggests an evolution in the au- 

 thor's mind. For instance, the ultra-basic 

 rocks were not separated as a distinct group in 

 Volume I., but now they appear as Division 6, 

 and the way in which factors of the quantita- 

 tive system are applied to give precision to the 

 new scheme is illustrated by the statement that 

 Division 6 embraces the rocks of Classes IV. 

 and V. of the quantitative system. Iddings's 

 first systematic division is actually by the 

 amount of normative salic and femic mole- 

 cules, into two groups, one corresponding to 

 Classes I., II. and III., and the other to 

 Classes IV. and V., of the quantitative system. 



Divisions 1-5 are bounded by sharp lines 

 determined by normative quartz, feldspar and 

 lenads, the same relation by which orders are 

 formed in Classes I., II. and III. in the quan- 

 titative system. The relative amount of mafie 

 (ferromagnesian) and felsic minerals is used 

 to make two subdivisions in each of these five 

 major divisions. 



The feldspathic rocks of divisions 2, 3 and 4 

 are each divided into three groups by the rela- 



