Reviews — Classijlcation of Rocks. 423 



kingdom, should be into — 1, Classes having at least one important 

 character in common ; 2, Into genera, having this and other leading 

 characters also common to the family ; 3, Into species, having, of 

 course, special and individual peculiarities. We find, however, that 

 the classification adopted by Dr. Cotta does not satisfy any of these 

 general principles of arrangement. For instance, in the 1st section 

 (p. 5) are placed together the oxides of Silicon and Aluminum. 

 Now, what character have these two substances in common, save 

 their combination with oxygen ? — an almost universal character in 

 minerals with which the geologist has to deal. 



Quartz and Corundum are not alike in any of the properties 

 usually considered important by mineralogists and chemists. Their 

 mode of occurrence — their chemical constitution^ — their crystalline 

 form — and, in fact, every other character, would separate them. 



The only character given for Amethyst (p. 6) is, that its colour is 

 violet, whereas it may be, and frequently is, yellow, or colourless ; 

 its real character being its right and left handed forms in the same 

 crystal, as shown long since by Sir David Brewster. 



Sodalite, Lapis-lazuli, Haiiyne, Nosean, Leucite, and Nepheline, 

 are all classed together (p. 14) ; we cannot see on what grounds. 



The Andalusite section (p. 34:) is made to include Tourmaliae,- 

 Topaz, and Lievrite. 



Under the heading of "Oxides of Elements of the Hydrogen 

 Group," (p. 60) we find the most wonderful oUa podrida in the 

 whole chapter. The group is divided into "A. Hydrous," and "B. 

 Anhydrous." What is an element of the Hj'-drogen group ? Dr. 

 Cotta implies that the classification he has adopted is a chemical one, 

 whereas we find, that, instead of containing the oxides of the chemist's 

 hydrogen group, which are Hydrogen, Potassium, Sodium, Lithium, 

 Silver, and some others, we have Spinel, Magnetite, Chromite, which 

 should have been, and are generally, put together as a separate 

 gToup of oxides [RO + E-a^a] placed with Hematite, a sesquioxide — 

 with Rutile and Cassiterite, oxides of di-atomic, or tetra-atomic 

 elements — with Pyrolusite, Hausmannite, and others equally incon- 

 gruous. Amongst these there is not one that can with any good 

 reason be placed within the Hydrogen group, which is essentially 

 mono-atomic, whereas the elements placed in it by Dr. Cotta are 

 di-, tri-, or tetra-atomic. He has omitted entirely from this group the 

 oxide of its type-element, to wit. Water or Ice, which is equally 

 interesting to the geologist and mineralogist, playing as it does such 

 an important part both on the surface and in the interior of our 

 globe. We find it, however, in the Miscellaneous division of 

 Part II. (p. 347), under the head of " Ice," but the notice of it is 

 rather meagre. 



The Augite (p. 16), Zeolite (p. 28), and Feldspar (p. 8) sections, 

 seem well arranged, and contrast advantageously with the foregoing 

 misalliances. 



^ Dr. Cotta's adoption of the formula Si O3 for Quartz brings it no nearer to 

 Corundum, which is a sesquioxide ; the modern sckool of chemists seem to consider 

 Silicon either di-atomic or tetra-atomic. 



