442 O. A. Derby—Occurrence of Gold in Brazil. 
may be compared to films of limonite deposited from an 
aqueous solution, or of mercury deposited from suspension in 
fatty matter. It seems impossible to regard these gold films as 
transported granules simply lodged against the surface. Their 
perfect adaptation to the surface on which they rest, their 
almost inappreciable thickness and their uniformity of struc- 
ture and of surface, whether large or smail, are against this 
view. In order to judge of the probability of this hypothesis 
I made the following experiment to determine the state in 
which fine gold occurs in the rocks of the region. A fragment 
of arsenopyrite assaying about five ounces to the ton, com- 
ing from about 500 meters below the surface in the Morro 
Velho mine, was dissolved in acid. The exceedingly minute 
specks of gold obtained in the residue are distinctly granular 
and crystalline. The smallest, an octahedral crystal, measures 
0-075 millimeters, while the largest, an irregular group of ecrys- 
tals, is 0°375 millimeters long. The fine gold of Sao Gongalo 
described in the following article, is of the same character. 
Such grains could not possibly be lodged against the surface so 
as to produce films of the appearance of those here described. 
The V-shaped streak of earthy limonite shows some pecu- 
liarities that seem to throw light on the process of introduction 
of the gold. Its outer margin is distinctly thickened, showing 
that the iron-bearing solution stood for some time upon it with 
a tendency to flow onward, causing a piling up of the liquid at 
the margin and a consequent thickening of the matter de- 
posited from it. As before remarked, the gold is more abund- 
ant on this streak than elsewhere on the specimen, and more 
abundant on the outer than on the inner half of it, as would 
naturally occur if it had been deposited from a solution which 
ad been for some reason, dammed back, so as to remain for 
some time on the place of the streak while it slowly evaporated, 
as the iron deposit proves to have actually occurred. On 
removing a little of the iron crust similar films of gold are 
seen resting on the botryoidal surface underneath the streak. 
These had evidently been formed before the accumulation of 
liquid from which the streak was deposited. Some of the 
gold films outside of the limits of the streak were also most 
probably deposited at the same time and in the same manner. 
Of these the greater number occur in a considerable patch, 
slightly discolored with iron, just below the angle of the V, as 
if the liquid which deposited gold on the streak had flowed 
over at this point, that of greatest accumulation, on the adjoin- 
ing surface, carrying a little iron with it. 
The facts above noted appear to be susceptible of the follow- 
ing interpretation as regards the course of events in the history 
of this specimen. In a narrow open slit of a quartz vein con- 
