150 J. O. Br miner — Geology of Fernando de Noronha. 



columnar structure either very plainly or partially developed. 

 The somewhat irregular columns lie for the most part horizon- 

 tally, as is well shown in the dyke exposed at the Horta do 

 Pico, southwest of the Peak. On the southeast side of the 

 Atalaia Grande the columns are nearly horizontal, with a tend- 

 ency to radiation from the center of the hill. Specimen No. 

 5 is from the nearly horizontal columns of phonolite exposed 

 on the southwest side of the Atalaia Grande. Similar columns 

 appear to form the whole south face of this peak, and are well 

 exposed near its base where they are washed by the surf. As 

 the peak is ascended the columns have a northward dip which 

 increases toward the summit, so that a north-south section 

 through the Atalaia Grande would expose the fan-like radiation 

 of the columns above mentioned. The preceding cut, from 

 a photograph of an exposure on the south side of Atalaia 

 Grande, fig. 2, illustrates this point. No. 40 is from the sum- 

 mit of Atalaia Grande. Nos. 41 and 42 are from a loose block 

 about 4x3x2 feet, found between Atalaia Grande and the 

 Morro do Meio, the hill lying immediately north of it. This 

 rock was not found in place on the island. It splits readily 

 into slabs very like the phonolites from the Pedra da Conceicao. 

 In the Sella a Gineta, fig. 4, the columns of phonolite are im- 

 perfect and vary somewhat in direction. Seen from Sao Jose 

 some of the columns curve from a horizontal position on the 

 left both upward and downward toward the right, radiating 

 from a horizontal axis. In the Peak too, the direction of the 

 columns varies in some cases as much as fifty degrees. The 

 lowest rocks of the Peak exposed in place are the irregular col- 

 umns upon its eastern side. The colums are here very nearly 

 vertical ; but higher up even upon this side, they twist and 

 bend to the northeast and thus form the overhanging projec- 

 tion which is so remarkable a feature of this great rock. 

 The curving of these columns prevents the falling of the 

 most picturesque part of the Peak.* On its western side the 

 columns stand at various angles with the meridian, and usually 

 at a high angle with the horizon. Their direction and 

 position, as well as the character of the rocks, leads to the 

 conclusion that the Peak is part of a great dyke, the only 

 remnants of which now exposed are the upper portions 

 of the Peak itself, and the columns at the Horta do Pico, a 

 short distance to the southwest. Specimens 51, 52 and 88 are 

 from the highest accessible part of the Peak on its southwest 



* Mr. Darwin calls attention to the disposition of phonolites to take on gro- 

 tesque shapes. (Geological Observations, second ed., pp. 97-8). On Fernando 

 these grotesque shapes of phonolite peaks are due to columnar structure when 

 the axes of the columns change their directions. For a discussion of the curving 

 and radiating of columns of igneous rocks see J. P. Iddings in Am. Jour. ScL, May, 

 1886, and Professor T. G. Bonney in Quart. Jour. Geol. Soc, vol. xxxii, 1876. 



