Williams — Petrography of Fernando de JSToronha. 181 



nor can the section be etched, except in the occasional 

 patches of glass, by the same reagent. Mr. Gill determined by 

 means of the Thoulet solution that all the powder of this rock, 

 except its heaviest components (magnetite, hornblende, titanite, 

 etc.), had a specific gravity between 2*557 and 2*155. All of 

 these tests, as well as the general character and structure of the 

 specimen seem to show conclusively that it contains no nephe- 

 line. 



No. 129, from a dyke cutting the rock last described, pos- 

 sesses a somewhat anomalous character. The lack of a com- 

 plete analysis, together with the altered condition of this speci- 

 men makes it difficult to define its exact position, but it 

 probably belongs somewhere between the trachytes and ande- 

 sites. A macroscopic examination is able to detect only a gray, 

 somewhat vesicular groundmass, enclosing occasional needles of 

 black hornblende. The vesicles are elongated in the direction 

 of movement in the mass before its solidification and are coated 

 with calcite and minute crystals of anal cite. 



The microscope discloses crystals of brown hornblende and 

 purplish gray augite imbedded in a fine network of long, lath- 

 shaped microliths. These latter when examined with a high 

 power show a very weak double refraction, suggesting melilite 

 or nepheline. Their habit, however, differs wholly from that 

 of these minerals, while their resistance to the action of strong 

 chlorhydric acid leaves little doubt that they are feldspar. 

 These little microliths are often quite completely changed to 

 calcite, and yet there is rarely visible the twinning striation 

 characteristic of the members of the plagioclase series. A spe- 

 cific gravity determination yielded no satisfactory results on 

 account of the alteration of the rock, the fineness of grain and 

 intimate admixture of magnetite. There was a continual fall 

 of the light-colored portions of the powder in the Thoulet 

 solution between the limits of andesine and sanidine, but the 

 presence of magnetite grains and zeolites prevented any con- 

 clusion being drawn from this fact. The only other original 

 constituent observed was biotite, in very minute but sharply 

 defined hexagonal plates. This is very abundant. The anal- 

 cite of the vesicles shows between crossed Nicols its anomalous 

 double refraction in great perfection. The percentage of silica 

 in this rock was determined, by Mr. John White, Jr., as 50*1. 



No. 10. Amphibole Trachyte, base of Atalaia Grande. — This 

 specimen presents a pale gray lithoidal groundmass similar to 

 the last, but the porphyritic constituents are here much more 

 abundant. These consist of glassy idiomorphic sauidine crys- 

 tals, black idiomorphic hornblende crystals, and rusty yellow 

 spots where some mineral no longer determinable has weath- 

 ered out. Under the microscope the sanidine appears quite 



