CHAPTER IV. 



THE CLASSIFICATION OP VOLCANIC EOCKS. 



Objects to be gained by a system of classification. — Artificial and natural systems. — The best system 

 represents -with accuracy the existing knowledge. — Progress is from the artificial to the natural 

 classifications. — All are evanescent and temporary. — Classification of volcanic rocks chiefly with 

 reference to physical properties. — Transitions to porphyritic rocks. — Correlations between physi- 

 cal properties. — Chemical composition. — Mineral ingredients. — Texture. — Density. — Fusibility. — 

 Wholly crystalline and partly crystalline textures. — Texture as correlated to geological age of 

 eruptions. — Not universally a true correlation. — Pre-Tertiary lavas common. — Von Cotta's view 

 adopted. — View tested by comparison with facts. — Magmas of all ages the same. — Texture due to 

 conditions of solidification.— Porphyritic texture. — Difficulty of definition. — No strict demarka- 

 tion between porphyries and lavas. — Crystalline rocks. — Signi6cancc of the wholly crystalline 

 texture. — The two original groups. — Acid and basic rocks. — Subdivision of each. — Andesite. — 

 Ehyolite. — The four major groups. — Conspectus of minerals characterizing the primary divisions. — 

 Ehyolites. — Trachytes. — Andesites. — Basalts. — General system. 



The objects to be gained by a good system of classification I hold to 

 be mainly two : first, accuracy of designation ; and, second, convenience of 

 treatment. In speaking of any natural object, it is desirable to indicate by 

 a single word as much as possible concerning the attributes and relations 

 of that object, and to avoid as far as possible all confusion with the attributes 

 and relations of other objects. In order to secure this accuracy and con- 

 venience it is necessary that a classification should be so constructed as to 

 express both the differences and community of attributes and relations. 

 "Where the differences of attributes between two or more objects are small 

 and the community of relations is nearl}^ complete, these objects are grouped 

 together as to most of their features, and separated only by small distinc- 

 tions, as varieties or species. AT here these differences are very great, and 

 the community very highly generalized, they are separated by much broader 

 divisions, as in orders or classes. When a category of objects is once clas- 

 sified and familiarized to the mind, the mention of any one of them will con- 

 vey not only an idea of the concrete object itself as an individual, but also 



