532 Progress in Science. [October, 
electro-magnet, with vibrating armature and conducting wires. The dial of the 
aneroid is formed of a disc of ebonite, 3-16ths of an inch in thickness. In this 
disc are insulated the steel index of the instrument and sixteen platinum points, 
forming terminals to the wires from the electro-magnets. To the index-hand 
is fastened one end of a thin strip of platinum foil. As the hand passes from 
1-1oth of a degree to another, it brings the platinum foil in contact with one 
of the platinum points; and, as the hand is one terminal of the battery, the 
bell corresponding to the platinum point brought into contact is rung, and 
continues ringing until the hand passes into contact with another point, or 
until the circuit is broken by a disconnecting switch. In the instrument con- 
structed for Captain Busk by Mr. Browning there are sixteen points, but only 
ten bells have been found necessary. The object of the inventor was to devise a 
simple and efficient apparatus which should indicate unerringly, on board 
ship, or to a person in charge of a lifeboat station, any important approaching 
alteration in the state of the atmosphere; and it is interesting to note the 
rapidity with which these changes occasionally suceeed each other. 
Prof. Edwin J. Houston contributes the following to the ‘ Journal of the 
Franklin Institute :’—‘* Having occasion recently to set up a large battery for 
experimental illustration of the properties of the voltaic arch, I noticed a fact 
which I believe has hitherto escaped observation. The battery consisted of 
about eighty half-gallon cells; fifty-five were Browning’s modification of the 
nitric acid battery of Grove. The negative element consists of sawed strips 
of very dense coke, the positive element of zinc, so arranged as to use both 
surfaces of the coke. The remaining cells were of the iron battery. When 
first set up, the arch between the carbon points measured fully 2 inches, while 
the flame frequently reached an equal distance above the upper carbon. The 
quantity of the current was very good—much better, in fact, than the size of 
the plates would have led me to expect. The phenomenon to which I would 
call attention is as follows :—‘* Wishing to show the well-known experiment 
of the rotation of the light bya magnet, I approached a compound bar magnet 
to the light, holding it with one end pointing diredtly to the arch, in a hori- 
zontal plane, equidistant between the carbon electrodes. When the nearest 
end of the magnet was 4 inches from the electrodes, the light was instantly 
extinguished. The regulator of the light which was employed is a form 
recently patented by Browning, of London. ‘The carbon points are kept at a 
constant distance from each other by the action of a small magnet, worked by 
the battery current. Though inapplicable to small batteries, with the current 
I employed it gave a light admirable for its steadiness. Thinking that the 
extinguishing of the light was produced by some cause other than the approach 
of the magnet, the experiment was repeatedly tried in a number of ways, 
until it was clearly shown that the cause could not be attributed to accident, 
but to the approach of the magnet. Though I have failed to find any pub- 
lished notice of this phenomenon, it seems probable that it may already have 
been observed, as the conditions of the experiment would be almost exa@ly 
reproduced whenever the rotation of the light of the voltaic arch by the mag- 
net was tried. Still it may be conceived that, though the necessary conditions 
for success in this experiment have often been nearly produced, they have sel- 
dom, if ever, been exactly produced; for it was noticed that in no case was 
the light extinguished unless the length of the arch was nearly as great as the 
tension of the electrode admitted ; that is, unless the electrodes were separated 
by nearly their maximum distance, consistent with the passage of the current. 
Were this condition not observed, in all cases, the approach of the magnet 
produced no other effe& than the rotation of the light, until it assumed a posi- 
tion in a vertical plane go° from a similar plane passing through the magnetic 
axis of the bar. Then, again, another necessary condition is that both the 
tension and the quantity of the current be of a strength greater than that of 
the current on which the experiment of rotation is generally tried. I have 
experimented with flames when these latter conditions were absent, and, al- 
though the rotation was observed, the extinguishing of the light was in no 
instance produced. The compound bar magnet employed is formed of three 
bars, held together by brass screws. It is 1 foot long, 1 inch broad, and 3 inch 
