518 Washington — Contributions to Sardinian Petrography: 



B. Trachyte (I. 5. (1)2. 3). Sennariolo, Monte Ferru. H. 

 S. Washington analyst. 



C. " Sanidine-plagioclase trachyte." Miniera, Monte Ferru. 

 C. Doelter analyst ; Denkschr. Ak. Wiss. Wien, xxxix, p. 51, 1878. 



D. " Sanidine-augite trachyte." Near Sennariolo, Monte 

 Ferru. C. Doelter analyst, op. cit., p. 54. 



E. Trachyte obsidian (I". 5. 1. 3.). Monte Rotaro, Ischia. 

 H. S. Washington analyst, this Journal, viii, p. 289, 1899. 



F. Phonolitic trachyte (I. 5. 2. 3"). Lusclade, Mont Dore, 

 Auvergne. Pisani analyst. Michel-Levy and Lacroix, C. R., 

 cxlviii, p. 1723, 1909. 



The two types are much alike, the only differences of note, 

 and these not very great, being in silica, lime and potash. 

 The preponderance, both actual and molecular, of ferric over 

 ferrous oxide may be noted. They resemble quite a number 

 of analyses of trachyte and syenite from other localities, two of 

 which are given for comparison. Doelter's analysis (C) of the 

 Cuglieri type much resembles mine (A) in its main features. 

 His analysis of the Sennariolo type differs markedly from 

 mine, his silica being much lower, and his iron oxide, magne- 

 sia, lime and soda higher. This is the more remarkable, as the 

 type is an unusually uniform one microscopically and both 

 specimens came from very nearly the same locality. 



The norms calculated from my two analyses are as follows : 



A B 



Q 2-04 2-52 



Or .. 34-47 42-26 



Ab .... 4035 40-87 



An 11-95 7-23 



C _.. .... 1-12 



Di 0-65 



Hy 1-20 



Wo 1-63 



Mt 0*93 0-23 



II 2-13 1-37 



Hm 2-56 2-40 



Ap 0-34 



These place both rocks in pulaskose, that from Cuglieri being 

 intermediate toward monzonase, symbol V. 5. 2. 3, and that 

 from Sennariolo being transitional toward phlegrose, symbol 

 I. 5. (1)2. 3. Microscopic examination alone, without chemical 

 analysis, would lead one surely to consider the Sennariolo type 

 as heralkalic. 



Mode. — The texture of these rocks is such that a satisfactory 

 measurement of the modes by Rosiwal's method is impractica- 



