694 The American Naturalist. [August, 
those that contain the smaller granules of iron. (Phil. Soc. Washing- 
ton, Bull. Vol. xii, pp. 39-52.) 
On the Separation and Study of the Heavy Accessories 
of Rocks.—For the past few years Prof. Orville Derby, of the Geo- 
logical Survey of Brazil, has employed the method of washing rock 
powder in water as the best means of isolating the accessory elements 
in rocks from the more abundant essential elements. His process dif- 
fers from that of Cordier and Thurach in that he uses the batéa or 
Brazilian miner’s pan instead of the glass or porcelain dishes of the 
laboratory. 
The knack of washing is readily acquired, and the whole operation, 
from the preliminary crushing to the mounting of a microscopic slide, 
can be performed in a few minutes. 
The delicacy of the process is well illustrated by the grouping of the 
minerals during the last stages of the washing. On throwing the sand 
out into a trail it will be found to be transversely streaked with differ- 
ent colors as the minerals arrange themselyes according to their spe- 
cific gravities. In a decomposed granite, for instance, the outer part 
will be white with quartz, then a reddish band of garnet, then a black 
band of titaniferous iron, and, finally (if magnetite be not present), a 
white band of zircon, oy white and yellow if monazite is also present. 
Prof. Derby’s observations incline him to think that the use of the 
batéa will prove valuable in work which is more strictly geological, for 
the following reasons : 
1st. Most of the prominent rock groups afford residues which are 
characteristic, either through the presence of accessories peculiar to 
each group, or by the relative abundance or peculiarities of form and 
structure of those that are common to several groups. 
2d. Many of the most common heavy accessories are practically 
indestructible, and can therefore be recovered in recognizable form 
eyen when the rocks or their debris are so completely altered that their 
original type is not otherwise recognizable. 
If rock types can be identified by their residues many rock masses 
can be taken into account which otherwise must be left out. 
For purposes of identification only material decomposed in situ 
should be washed. 
Prof. Derby also suggests that this method of study of the heavy 
accessories of rocks may be of value in investigating those rocks which 
have suffered alteration by metamorphism. 
