SCIENTIFIC SUMMARY. 
435 
they contain ; and these vibrations being communicated to the air in the 
cavities, under the most favourable conditions for producing sound, the result 
is the loud noise which is caused when any large mass of sand is set in 
motion. We have, in fact, millions upon millions of resonant cavities, each 
giving out sound which may well swell up to resemble a peal of thunder, 
with which it has been compared; and the comparison — I know from others 
who have heard it — is not exaggerated. The effect of rain in preventing the 
sound is owing to the cavities in the sand becoming filled with water, and 
thus rendered incapable of originating vibrations.” 
PHYSICS. 
A New Form of Mountain Barometer has been described by Mr. Emmons 
in “Appalachia,” June 1876, the first number of a new American journal 
devoted to mountaineering pursuits. In the ordinary mountain barometer 
we must carry about with us a tube of nearly a yard in length, but 
Macneill’s instrument need not be more than half this length. The tube is 
open at both ends, and the lower end passes by an air-tight connection 
through the top of a cylinder, which opens only in a tube below, connecting 
it with the mercury cistern. The bottom of this cistern is of soft leather, 
resting upon the end of a vertical screw, as in the Fortin barometer, and the 
whole cistern, with its mercury, may be separated from the tube and carried 
independently, the liquid being retained by a stop-cock. To use the instru- 
ment, it is hung vertically, and the screw is turned up, forcing the mercury 
into the cylinder above. Then the liquid rises to the level of the bottom of 
the open tube, the air above it is confined in the cylinder, and is then under 
the atmospheric pressure. Now, if we continue to force up the mercury we 
compress the air in the cylinder, and the liquid rises to a corresponding 
height in the tube. This compression is continued in all measurements, 
until the mercury in the cistern rises to a certain fixed point. The corre- 
sponding height of the mercury column in the tube, as read upon its "scale, 
will be greater or less, according as we are reading at the sea level or upon 
a mountain, and if the scale has been previously graduated by comparison 
with an ordinary mercurial barometer at different pressures, the atmospheric 
pressure may be read directly from it. The error in reading this instrument 
is greater than in the Fortin or other barometers, and it is also liable to 
error from some other sources ; but it is an extremely convenient form for 
transportation, since the mercury may be carried separately from the tube, 
and the whole instrument is hardly more than half the length of the shortest 
syphon barometer of the same range. 
New Experiments with the Radiometer have been conducted by M. A. 
Ledieu, and are reported as follows in the “ Comptes Rendus ” (June 12) 
and “Chemical News” of July 14: — The radiometer was found to continue 
revolving when submitted exclusively to a pencil of luminous rays falling 
parallel to its axis. The author, however, does not draw the conclusion to 
which a superficial and systematic examination of this result might seem to 
lead. The experiment performed by M. Salleron at the suggestion of the 
ff2 
