50 
SEA SAND. 
edges. As more grains follow, they pile up in the middle 
making a low volcano shaped cone, until the pattern is spoilt 
by the middle parts beginning to slide and forming a cone of 
constant angle as often happens with a cinder cone. 
Some of the mathematical problems suggested by the sand 
grains are similar to those which occur in the theory of mole¬ 
cules. The diagram of the bullet marks on the target is 
familiar in the Kinetic theory of gases, although in that case 
the discussion concerns the distribution of diverse velocities 
among the molecules. More recently Prof. Osborne Reynolds 
has used the properties of sand to suggest that possibly not 
only matter ibut also ether behaves as though it had a 
granular structure. 
The size of the sand grains determines the size of the spaces 
between them; and many of the properties of sand depend 
upon these intervals. We have attempted to get some measure 
of these interspaces by finding how much water they will hold. 
A dry measure of ioo cubic centimetres of sand is somewhat 
indefinite unless the sand be well shaken together ; for under 
judicious shaking the volume will be diminished as the 
particles pack closer together. A similar closer packing 
occurs with spherical lead shot. 
A 200 cc. measuring jar was half-filled with water up to the 
ioo cc. mark, and a measured ioo cc. of sand poured in. If 
there were no interspaces the total volume would have been 
200 cc., but the total fell short of this by some 30 to 40 cc., 
showing that about j, of the dry sand had been air-space. 
This fraction is not the same for different kinds of sand. 
Can wejpredict whether it will depend on the size or shape of 
the grains ? Will small shot or large shot allow 7 most air¬ 
space ? or round shot or angular sand ? and what about a 
mixture of many shapes and many sizes ? 
0000 
0000 
0000 
0000 
000 
0 0 0 
0000 
0000 
0000 
0000 
000 
000 
Fig. 1 . 
IT i 0 * o 
Fig- 3- 
