ON HORNBLENDE ANDESITES FROM THE NEW VOLCANO ON BOGOSLOV ISLAND IN 



BERING SEA. 



BY GEOBGE P. MERRILL. 



The rocks described below were received from Capt. M. A. Healy, by whom they were collected 

 and donated to the National Museum. On account of the interest just now attached to the locality 

 they seem worthy of a special description. It is well to remark in the beginning that none of the 

 samples received show freshly fractured surfaces, but are in the form of irregular blocks with 

 their corners broken and rounded. They were accompanied and covered with a fine sand and 

 dust of the same mineral nature as the rocks themselves, but stained by sulphur and iron oxides. 

 Evin in the absence of definite information on the subject, it seems safe to infer that they are 

 simply ejected volcanic blocks, and not from recent lava flows, none of which have as yet been 

 reported. 



Two varieties of the rock were received, one consisting of a light gray, slightly purplish, fine 

 grained and porous groundmass, in which small glassy feldspars and dark brown and green horn- 

 blende-like crystals are readily distinguishable by the naked eye. The texture is quite uniform, 

 the brown hornblende being the more variable constituents, in one case a single crystal nearly 

 half an inch in diameter being observed. The rock is rough to the touch and somewhat friable. 

 Under the microscope it is fouiul to consist of a light gray groundmass, in which are embedded 

 deep reddish brown, strongly dichroic hornblendes, light green augites, and numerous crystals of 

 a plagioclase feldspar, together with scattering grains of iron ore. The hornblendes are usually 

 in irregular crystals, though an occasional quite perfect basal section was observed which showed 

 a preponderance of the prismatic faces. The crystals are often elongated in the direction of their 

 vertical axes, and a portion of them show the dark borders so coniniordy seen in the hornblende 

 of andesites. The-augite is light green in color, and at first ghmce might readily be mistaken for 

 a green variety of hornblende. Its cleavage and optical properties are, however, unmistakably 

 those of augite. 



In form the plagioclases are short and thick, showing but few twinning striations, sometimes 

 none at all. A portion of them are clear and pellucid, while others are clouded through the pres- 

 ence of numerous cavities and impurities. In many cases the outer portion of a crystal is clear, 

 while the interior is clouded, or again both outer and interior portions may be clear while there 

 exists an intermediate zone full of cavities. In addition to these the plagioclases contain numer- 

 ous inclosures of a yellowish glass, which often bears a bubble, and hornblende and augite parti- 

 cles. A number of short and thick, clear, glassy feldspars are present, which show no sign of 

 twinning, and which appear from their optical properties to be sanidin. Both sanidin (?) and 

 plagioclase show at times a very evident zonary structure. 



Apatite occurs but sparingly and in minute colorless crystals, showing but slight trace of the 

 dusky interiors so often seen in the apatites of this class of rocks. The magnetite is, as a rule, in 

 but poorly defined crystals. 



The base proper of the rock consists of an aggregate of minute colorless microlites* and grains 

 of opacite; there is also present a very weakly doubly refracting, colorless, interstitial substance, 

 which, under a power of 300 diameters is seen to be composed of rounded and irregular imbricated 



*In a preliminary note on these rocks, published in Science of December 12, 1884, the base was stated by mistake 



to be miorofeUitic. It should have read micyi'oUtic. 



45 



