February 27, 1903.] 



SCIENCE. 



343 



assuLva. If for any purpose we wish to cal- 

 culate the percentage mineral composition of 

 the rock, giving: the exact proportion of min- 

 erals actually present, full exijlanations as to 

 the method to be followed are also given, this 

 constituting the mode of the rock, or the man- 

 ner in which the chemical elements have 

 actually arranged themselves. As a general 

 rule, the norm and the mode of a rock agree 

 closely. 



Xo classification, however, which requires 

 a chemical analysis of a rock before the posi- 

 tion of the rock can be determined and a 

 name given to it, would be susceptible of gen- 

 eral use. Consequently a method is indicated 

 by which it is possible to determine the chem- 

 ical composition of a rock without the aid of 

 such an analysis. Eosival has recently shown 

 that if a few thin sections of a rock are taken, 

 the relative proportions of the various min- 

 erals constituting the rock may be determined 

 by measuring under the microscope the di- 

 ameters of each crystal in lines running arbi- 

 trarily across the thin sections in question, 

 care being taken to measure a distance at 

 least 100 times as great as the average di- 

 ameter of the constituent grains. The values 

 obtained will correspond to those of the vol- 

 umes of the several minerals present. The 

 relative weights of the several minerals may 

 be deduced from these volumes by multiply- 

 ing each by the specific gravity of the mineral 

 and reducing the whole to 100 parts. 



The approximate chemical composition of 

 the several minerals can be deteiinined from 

 the known composition of these species in 

 similar rocks, and from these data the com- 

 position of the rock as a whole can be easily 

 calculated. From this, in its turn, the norm 

 may be obtained. 



In the case of glassy rocks or those con- 

 taining a large amount of imindividualized 

 material, a chemical analysis is necessary, 

 just as it is for that matter in many cases in 

 the system of classification now employed. 



The norminative mineral composition and 

 the chemical composition of the rock being 

 thus ascertained, its position in the classifica- 

 tion can be readily determined. For this 

 purpose the rock-making minerals are divided 



into two groups, namely, those which are char- 

 acterized by a high content in silica, alumin- 

 ium and alkalis, and those characterized by 

 a high content of iron and magnesia. The 

 first group is known nuiemonically as the 

 salic (silica-alumina) group and the second 

 as the femic (ferro-magnesian) group. On 

 the relative proportion of the minerals of 

 these two groups present, rocks are divided 

 into five classes, according to whether one or 

 other of these groups is extremely abundant 

 or merely dominant, or whether the minerals 

 of the two groups are present in about equal 

 proportions. These five classes are thus char- 

 acterized (commencing with the most salic) 

 as the persalane, dosalane, salfemane, dofe- 

 mane and perfemane. These classes are sub- 

 divided into orders according to the relative 

 proportions of minerals forming the predom- 

 inant group in each case. Thus in the pre- 

 ponderatingly salic classes the order will be 

 based on the relative amount of quartz, feld- 

 spars and feldspathoids. The orders in their 

 turn are subdivided into rangs (an archaic 

 equivalent of ranks used to avoid confusion 

 with this latter term), on the ground of the 

 chemical character of the bases in the min- 

 erals of the preponderant group in each case; 

 thus, if these were feldspathic, the fivefold 

 division would be made according to the pro- 

 portion of alkalis to lime in the feldspars. 

 The lowest division, known as a grad (an 

 archaic form of grade), is based on the rela- 

 tive amounts of the minerals composing the 

 subordinate group in the rocks. In addition 

 to these, further subdivisions are provided for, 

 when necessary, by subclasses, suborders, siih- 

 rangs and subgrads. 



The system demands an entirely new no- 

 menclature; in fact any attempt to adapt the 

 old nomenclature to the new system would re- 

 sult in the direst confusion. A new nomen- 

 clature in its entirety has accordingly been 

 elaborated, but, being based upon a definite 

 plan, is easily grasped after a little practice. 

 Each name consists of a root derived from 

 some geographical name, the name of some 

 locality where the rock in question is typically 

 developed, the localities being chosen impar- 

 tially from all countries, thus giving an inter- 



