P.G.H. Boswell— Quantitative Methods in Stratigraphy. 109 
of quartz. It is clear that each grade obtained by elutriation will 
contain a proportion of heavy grains which from their diameter ought 
to belong to a smaller grade. If the percentage of material greater 
than 2°8 in density were a considerable one, the method of grading 
would be vitiated. Apart from the occasional cases where quantities 
of such authigenic minerals as limonite, pyrite, and pyrrhotite, etc., 
are introduced (and these may be extracted early in the operations), 
the heavy crop reaches only 4 per cent by weight in special cases. 
The average found by the writer after numerous separations of rocks 
of all ages is about ‘6 per cent. It is often, especially in coarser 
sands and sandstones, much less. Fine sands consisting largely of 
grains *2 to ‘05mm. diameter, such as many of the samples of 
Bagshot Sands, appear to yield the largest residue (up to 4 per cent).” 
The earliest systematic work in this country on heavy detrital 
minerals appears to have been carried out by Allan Dick, and he 
was fortunate in choosing for his work Bagshot Sand from 
Hampstead Heath, which yields a large crop. Sands of diameter 
2 to ‘5mm. (e.g. Red Crag, etc.) yield a much smaller proportion 
of heavy minerals. 
Ili. Tue Possrprnitres anp Limrrations oF THE MerHops. 
A Imowledge of the mechanical analyses of all the British 
incoherent sediments is economically valuable as well as of con- 
siderable geological importance. Our resources are not at present 
accurately known, but soil-analysts, authorities on water-supply and 
filtration, brick and pottery manufacturers, glass manufacturers, and 
workers in various branches of the engineering trades, particularly 
in the foundries, agree upon the importance of the data. From the 
purely geological point of view, many interesting deductions can be 
made regarding conditions of deposit, velocity of rivers and currents, 
and direction of drainage. We need, however, much more experi- 
* mental work like that inaugurated in this country by Forbes, Sorby, 
and others, and carried on abroad, particularly in Germany and the 
United States,> in properly equipped laboratories. The experimental 
work is largely synthetic, and seeks to build up deposits under certain 
known conditions; the corresponding analytical work upon geological 
deposits has rather tended to lag behind the synthetic—a result not 
expected in petrology. When sufficient data have accumulated it 
should be possible to devise schemes for graphical representation 
of sedimentary rocks, similar to those in use for igneous rocks. 
Diagrams might be used upon maps, and being placed at certain 
points over an outcrop, yield at a glance information as to heteropic 
or isopic formations, lithological changes, presence of shore-lines, etc. 
But not only should deposits actually mapped at the surface be so 
treated. Specimens from wells, water- and trial-borings should be 
carefully preserved and subjected to analysis. It is a matter of 
' R. H. Richards, Ore-Dressing, 1906, tables in Appendix. 
2 Exclusive of such local sands, etc., such as those bordering the granite 
masses of Devon and Cornwall. These sands may be full of tourmaline, etc. 
® See, for example, the recently published Professional Paper 86, U.S.G.S., 
‘‘The Transportation of Débris by Running Water ’’ (G. K. Gilbert). 
