Prof. T. O. Bonney — On Dana's Classification of Rocks. 201 



rock as perfectly defined when we say it consists of " plagioclase 

 and augite," and so forth ; but we not seldom find that we must be 

 satisfied with this ; because even Professor Dana himself could give 

 us no better definition. Let us grant that there be a typical albite, 

 oligoclase, labradorite, and anorthite ; that there may be rocks in 

 which each one of these felspars alone occurs ; still I maintain that 

 they have a frequent habit of intergrowth, as in perthite, bytownite, 

 microcline with albite, etc., and that in many cases they thus occur 

 in rocks. Examination also of the analyses given by Professor 

 Dana himself 1 shows that even among the type felspars there is 

 great variety ; and that when a felspar is a rock constituent we have 

 no security against variation or intermixture of species ; while in 

 many cases the felspars occurring in rocks are so minute that it 

 would be impossible for the chemist to separate them for analysis. 

 Analysis of the rock as a whole often fails to give the desired 

 response, because the chemical constituents which it reveals may 

 enter into the composition of more than one of the minerals already 

 known to be present ; so that the investigator is in the position of a 

 mathematician who is asked to solve an indeterminate equation. 

 Under certain circumstances we can, indeed, feel sure that the 

 plagioclase is either albite or oligoclase or both, under others that it 

 is labradorite or anorthite or both ; and it would perhaps be expe- 

 dient to add after the term " doubtless a or & or a mixture of them," 

 but we may be pardoned for avoiding the monotony of this constant 

 repetition, and leaving this to be supplied by the intelligent reader. 



Our position then, as workers with the microscope, is that while 

 thankfully acknowledging the aid of chemical analysis, we do not 

 feel bound to spend much time or money in determining the precise 

 character of the felspar in every case, because we know from experi- 

 ence that not seldom the oracle gives an ambiguous response, or at 

 best leaves us very nearly where it found us. 



Accordingly we use the term plagioclase 2 (with the understood 

 limitation) as the most definite permitted by the present state of 

 science. It may be that ultimately we shall be able to form a rock 

 group correspondent with each of the species of plagioclastic felspar ; 

 but I doubt it, and expect that we shall not do more than separate 

 those containing albite and oligoclase from those containing labra- 

 dorite and anorthite ; though even then it is not impossible that the 

 second and third of these may sometimes be associated. 



My most essential difference, however, from Professor Dana relates 

 to his proposed classification of rocks. In this he passes over the 

 question of the origin of the rock, by adopting a purely chemical or 

 mineralogical basis, and grouping metamorphic clastic with true igne- 

 ous rocks. He also justifies this method of classification by pointing 

 out that true massive crystalline rocks need not necessarily be igneous, 



1 A System of Mineralogy, pp. 337-361 (ed. 1868). 



2 It is no doubt inconvenient that the species microcline has heen discovered since 

 the adoption of the term plagioclase ; but we may avoid this difficulty by agreeing 

 that the term plagioclase shall be used as a symbol for the group of soda and lime 

 felspars — the character of the other species being so exceptional, and its relations to 

 orthoclase being in most respects close. 



