386 SCIENCE, 
of the vastly greater constructions which would be 
needed in any attempt to control floods. There is 
special insecurity necessarily inhering in the founda- 
tions of hydraulic work such as this, constructed in 
the bed and banks of running streams. Besides 
this, mention may be made of the interference which 
such reservoirs would cause in vested interests, such 
as mills, factories, railroads, canals, and rafting. 
The latter portion of Col. Merrill’s letter advises 
the city of Cincinnati to appoint a commission to 
define the lines of the river-front for high and low 
water, and to make it the duty of some public officer 
to prosecute in case of infringement on the free 
waterway, so that there may be no future obstruc- 
tion of the river-channel. He shows that to the 
present time there has been no perceptible obstruc- 
tion at this point. He further advises that the lower 
part of the city, next to the river, be filled by con- 
tinuing the present slope of the river-banks upward 
to high-water mark; and then that those squares of 
the city which stand on the slope be devoted to busi- 
ness alone, and be so solidly built as not to be seri- 
ously damaged by an occasional flood, while the 
houses of the laboring population be removed to other 
parts of the city. 
In conclusion, Col. Merrill says, in reply to the ques- 
tion which has frequently been put to him as to what 
the government is going to do to try to stop these 
floods, that, if the government be guided by his ad- 
vice in this matter, it will do nothing, as the undoubt- 
ed cause of the flood was the excessive fall of rain 
and snow; and no means of controlling this has yet 
been discovered. 
AN EXPLANATION OF HALL’S PHE- 
NOMENON.} 
Mr. E. H. HALUW’s original experiment was as fol- 
lows. A strip of gold-leaf was cemented to a plate 
of glass, and placed between the poles of an electro- 
magnet, the plane of the glass being perpendicular to 
the magnetic lines of force. The current derived 
from a Bunsen cell was passed longitudinally through 
the gold; and, before the electro-magnet was excited, 
two equipotential points were found by trial near 
opposite edges of the gold-leaf, and about midway 
between the ends. When these points were connected 
with a galvanometer, there was, of course, no deflec- 
tion. A current from a powerful battery being passed 
through the coils of the magnet, it was found that a 
galvanometer-deflection occurred, indicating a differ- 
ence of potential between the two points, the direc- 
tion of the current across the gold-leaf being opposite 
to that in which the gold-leaf itself would have moved 
across the lines of force, had it been free to do so. 
On reversing the polarity of the magnet, the direc- 
tion of the transverse electromotive force was re- 
versed; and, when the magnet was demagnetized, the 
two points reverted to their original equipotential 
condition. Subsequent experiments showed that the 
direction of the effect differed according to the metal 
1 Abstract of a paper read at the meeting of the Royal society, 
Feb. 21, 1884, by SHELFORD BIDWELL, M.A., LL.B. 
used. This effect was attributed by Mr. Hall to the 
direct action of the magnet on the current. 
Mr. Bidwell claims that Hall’s phenomenon might 
be completely explained by the joint action of me- 
chanical strain and certain thermo-electric effects. 
The, strain is produced by electro-magnetic action. 
It will be convenient to refer to the metallic plate or 
strip as if it were an ordinary map, the two shorter 
sides being called respectively west and east, and the 
two longer, north and south. Let the south pole of 
an electro-magnet be supposed to be beneath the 
strip, and let the strip be traversed by a current pass- 
ing through it in a direction from west to east: then 
the strip will tend to move across the lines of force 
in the direction from south to north. Since, however, © 
it is not free to move bodily from its position, it will 
be strained; and the nature of the strain will be some- 
what similar to that undergone by a horizontal beam 
of wood which is rigidly fixed at its two ends, and 
supports a weight at the middle. Imagine the strip 
to be divided into two equal parts by a straight line 
joining the middle points of the west and east sides: 
then in the upper or northern division the middle 
district will be stretched, and the eastern and west- 
ern districts will be compressed; while in the lower 
division the middle part will be compressed, and the 
two ends will be stretched. If, now, a current is 
passing through the plate from west to east, the por- 
tion of the current which traverses the northern 
division will cross first from a district which is com- 
pressed to one which is stretched, and then from a 
district which is stretched to one which is compressed ; 
while in the southern division the converse will be 
the case. And here the thermo-electric effects above 
referred to come into play. 
Sir William Thomson, in 1856, announced the fact 
that a stretched copper wire is thermo-electrically 
positive to an unstretched wire of the same metal, 
while a stretched iron wire is negative to an un- 
stretched iron wire. From this it might be inferred, 
as Sir William Thomson remarks, that a free copper 
wire is positive to a longitudinally compressed copper 
wire, and that a free iron wire is negative to a lon- 
gitudinally compressed iron wire; and experiment 
shows this to be the case. A fortiori, therefore, a 
stretched copper wire is thermo-electrically positive 
to a compressed copper wire, and a stretched iron 
wire is negative to a compressed iron wire. If, there- 
fore, a current is passed from a stretched portion of a 
wire to a compressed portion, heat will (according to 
the laws of the Peltier effect) be absorbed at the junc- 
tion if the metal is copper, and will be developed at 
the junction if the metal is iron. In passing from 
compressed to stretched portions, the converse effects 
will occur. 
It follows from the above considerations, that, if the 
metal plate (which is subjected to a stress from south 
to north, and is traversed by a current from west to 
east) be of copper, heat will be developed in the west- 
ern half of the northern division, and absorbed in the 
eastern half; while heat will be absorbed in the west- 
ern half of the southern division, and developed in the 
eastern half. But the resistance of a metal increases 
iad) 
: 
[Vor IIL, No, 607 
