head-water territory which is delivering the main mass of sediment, and there to locate a vein in place. The problem is complicated by multiplicity of tributaries and by large size of the drainage areas. In such cases careful panning and testing of the gravels at frequent intervals may show which of several tributaries are contributing most of the values, and thus may further localize the area of search. Many important mining districts, including Butte, Bisbee, the Mother Lode region of southern California, the diamond fields of Africa, and others, have been found by tracing up placers in this manner. In the case of an older placer deposit, where the topography and drainage have been much altered since its formation, or where the deposit has been covered by later sediments, the problem is of course much more difficult.
Much less than a commercially valuable placer deposit in unconsolidated surface rocks may start a search for the mother lode. A single fragment of ore in the "wash" naturally directs attention up the slope, and the repetition of fragments in a certain direction may lead unerringly to the source. The fragments may not even in themselves carry value, but may consist of detrital material from the leached outcrop—such as iron or manganese oxides, which, because of their red or black color, stand out conspicuously in the rock débris.
In the Lake Superior region large angular fragments of iron ore or iron formation in the glacial drift immediately raise question as to source. If the fragments are rounded and small, they usually indicate a very distant source. The general direction of glacial movement is known in most places, and by tracing up the fragments in this direction the outcrop may be found; or the chain of fragments may be traced to a point where they stop, which point may serve to locate the parent bedrock carrying the ore body, even though it does not outcrop.
An interesting suggestion was made some years ago with reference to the diamonds found sporadically in the terminal moraines in Wisconsin and other mid-west states. The diamonds are of such size and quality as to indicate surely the existence of a real diamond field somewhere to the north. The locations of these diamond finds were platted on a glacial map, and lines were projected in a general northerly direction along the known lines of the glacial movement. It was found that these lines converged at a point near Hudson's Bay. The data were too meager and the base line too short for this long projection, and the indicated source of the diamonds can be regarded as the merest speculation. However, with the finding of additional diamonds in the drift, as seems very likely, the refinement of this method might conceivably bring results in time.
The Use of Magnetic Surveys in Tracing Mineral Ledges
Magnetic surveys are often useful in tracing iron-bearing rocks beneath the surface, in the discovery of outcrops of such rocks, and in working out their lines of connection. This method is in general use for the crystalline iron ores in the Lake Superior region, Canada, the Adirondacks, and elsewhere in the glaciated portions of the United States. It is not so useful for the brown ores and the Clinton ores of the southeastern United States, which are only slightly magnetic and can be commonly located by other methods.
Where the ore is strongly magnetic, and is associated with other rocks which are non-magnetic, the nature of the magnetic field determined by a surface survey with vertical and horizontal needles may tell something about the shape and size of the ore body. Commonly, however, magnetic ores are associated with leaner magnetic rocks,—with the result that the magnetic survey, unless it happens to lead to an outcrop of ore, indicates only the general area through which underground exploration might be warranted. In the hematitic iron ores of Lake Superior, magnetism is less pronounced than in the magnetites; and in the soft