106 SNAKE RIVER PLAINS OF IDAHO. [btjll.199. 
was a strong flow, although smaller than the ones just referred to as 
being of decidedly recent date, and presents remarkable examples of 
large swells and stream-like pahoehoe ridges crossing one another. 
The lava, although bare and still fresh in general appearance, has 
evidently lost the bloom of youth, and its surface shows some indica- 
tions of weathering. The most interesting feature in this connection 
is the presence on the surface of the smooth pahoehoe swells of a thin, 
deep-blue incrustation or film of the nature of ' ' desert varnish. 7, « The 
thin films designated by this term occur on rocks which are exposed 
to an arid climate, and seem to be due to the solution by absorbed 
water of some of the material forming them, and to its transfer, 
through the action of capillarity as surface drying takes place, to the 
surface, where it is precipitated. The examples of desert varnish 
described by Gilbert and Diller are of various shades of brown deep- 
ening to black, and are due to a film of iron and manganous oxides. In 
the present instance the surface film is a strong, deep blue, closely 
similar to earthenware colored with cobalt. The color is a decided 
cobalt blue, but, as chemical tests show, is not due to that metal. 
Attempts to separate enough of the film to serve for analysis have 
failed, and it may be a physical rather than a chemical phenomenon. 
That the color is of the nature of a desert varnish and not due to 
oxidation as the lava cooled, in a way similar to the changes occurring 
on the surface of steel and other metals when highly heated and allowed 
to cool, is seemingly shown by its presence on exposed surfaces and the 
upper portions of the sides of fissures and its absence on the walls 
of cavities unless they were broken at about the time the lava hard- 
ened. The color is deepest near the borders of crevices and becomes 
lighter on the surfaces between the cracks. In many instances long, 
sinuous, stream-like ridges emerge indefinitely on the surface of the 
lava stream, become prominent for a distance of perhaps 10 to 30 feet 
and then disappear beneath other ridges or swells formed by similar 
local outbreaks of viscous lava. These ridges are covered with a net- 
work of small shrinkage cracks, the borders of which are deep blue and 
the central parts of the interspaces light blue or gray. The appear- 
ance of these tessellated serpent-like forms is frequently striking. 
Their comparison to scaled reptiles, although fanciful, is suggestive 
of their appearance. In my notes I have termed the lava stream bear- 
ing these unique forms the Blue Dragon lava flow. 
The opposite neighbor of the Blue D-agon lava flow, or the third 
recent lava stream on the north side of the Cinder Buttes, counting 
from the west, is not as extensive as the other streams which went 
northward, although it presents a typical pahoehoe surface fully half 
a mile in length. This stream, although still young, bears evidence of 
«Aterm proposed by G. K. Gilbert, and more fully explained by J. S. Diller in The Educational 
Series of rock specimens, Bull. U. S. Geol. Survey No. 150, Washington, 1898, pp. 389-391. 
