A PERIDOTITE DIKE IN PENNSYLVANIA COAL MEASURES 5 1 7 



alteration products are themselves in almost all cases so pale 

 in tint that one would expect a molecule near forsterite. In 

 but two slides were its secondary minerals noticeably green, 

 and then they were undoubtedly chlorite. The alteration has 

 gone so far or has taken such a course as not to show serpentine 

 veinlets in marked development, but by careful search both 

 fibrous chrysotile and scaly antigorite have been detected. 

 The identity of the olivine is in consequence a little less apparent 

 than is often the case. Abnormally large and corroded or 

 rounded phenocrysts of olivine are also present up to two or three 

 cm. in diameter, as was mentioned above. They are altered 

 like the rest but have fresh nuclei. To the unaided eye, they 

 resemble pyroxene more than olivine. 



Magnetite is quite richly distributed, and the grains may reach 

 a size several millimeters in diameter. Much of it is shown by 

 the alteration product to be titaniferous. Perofskite is abundant 

 in small, highly refracting, brown grains. Apatite is present. 

 One garnet with a kelyphite rim has been detected, and pyriteds 

 occasional. Almost all accurate traces of the ground mass, 

 whatever it was, have given way to calcite and dolomite, which 

 are richly present in the slides. Careful search has failed to 

 reveal the melilite which one would suspect to be present, from 

 the abundant calcite in the rock, nor can any other feldspathic 

 mineral be identified. In one instance isotropic material was 

 found, which was perhaps analcite, and through it was distrib- 

 uted very small stocky prisms which may once have been angites. 



An analysis kindly made by Miss M. W. Adams yielded the 

 following results: 



Si0 2 28.83% Na 2 75 



Ti0 2 5.67 K 2 1.31 



A1 2 3 2.94 H 2 + 3.96 



Fe 2 3 3.60 H 2 0- 0.83 



FeO 5. 13 CO, n.64 



MgO 24.31 P 2 O s ■ Q-77 



CaO 11.24 Total 100.98 



It is possible to recast this analysis only in an approximate 

 way, but by several assumptions a result can be reached which 



