SEA SAND. 
57 
with him a photographic camera and a supply of potassium 
ferrocyanide. 
Brown faded writing still contains the iron of the ink, and 
if it is brushed over with a weak solution of this potassium 
ferrocyanide, the letters appear again distinct and blue. And 
so examining this solution of sand by adding a few drops of 
potassium ferrocyanide, a bright Prussian blue colour appears 
which we regard as proof of the presence of iron. The yellow 
sands generally contain iron, staining the grains of quartz, and 
it appears to be this yellow iron compound which changes to 
pink when Scarboro’ sand is strongly heated. Presumably the 
change is by loss of water from some hydrate of iron to ferric 
oxide. The green colour of common glass is due to ferrous 
oxide derived from sand containing iron, and it is to avoid iron 
that makers of the best glass take such pains to secure a pure 
white sand. 
There are other metals besides iron for whose presence we 
have not yet searched. Nor has our microscopical examination 
of sand gone much further than admiration of the magnificently 
pqlished appearance of the grains when seen by reflected light. 
A further interesting inquiry would be to use certain very dense 
liquids to sort out the grains of sand according to their specific 
gravities. 
We have just noticed the influence of shape on the rate at 
which sand falls through water. The thin flat fragments of 
shell and the little flashing flakes of mica fall slower than the 
rounder grains, and therefore come to lie on the surface. In 
many sandstone rocks the mica flakes lie in planes, and it is 
along these planes that the rock splits most easily. 
The grains may be roughly sorted by shape, by shaking 
them on a sloping glassplate when the rounder ones tend to 
run off first. It is, of course, only a rough sort of artificial 
selection process, but if those that run off the plate are put 
through the process again and again, we obtain at last a small 
quantity of sand so round and so mobile as to run about almost 
like drops of mercury. 
A round sand grain will probably run before the wind better 
than a flat one, and I suspect that sandhill sand may prove of 
rounder quality than seashore sand. This is certainly true of 
two samples gathered from the Northumberland coast, a little 
north of Newbiggin-by-the-Sea; and Calais sand (which I 
suppose to be from the French sandhills) lacks the flat shell 
E 
