308 British Association for the Advancement of Science. 



wire Q, and with the zinc plate of the battery by the wire P ; and 

 into the cups c" and c n \ connected with each other by the wire X, 

 and with the copper of the battery by the wire D. rv rv and rv' rv' are 

 connected with the cups V and R by the wires K and N, and with 

 the poles of an electro-magnet M by the wires S and T. The wires 

 K and N rise in the cups of mercury V and R, but do not leave the 

 mercury. Let us suppose, as in the figure, the wires rv w and rv' rv' 

 to dip into the cups c c" : we shall trace the electricity. It flows 

 from the copper of the battery F along D to c'", along X to c", along 

 N to R, along T to S, where it enters one pole of the magnet M. 

 Now, let the wires rv rv and rv' rv' have turned a little on their axis 

 A B, so as to dip respectively into c' and c"'. The electricity flows 

 from F along D to c"', along rv re, crossing the axis A B, along K, 

 to V, along S, and enters the magnet at N, before it entered at S ; 

 hence the current is inverted and the poles reversed. 



The machine can be stopped or set in motion in a moment by lift- 

 ing or replacing any of the wires forming the galvanic circuit. 

 If its motion be interrupted its power is not wasted as in other ma- 

 chinery, but is accumulating ; so that when it again works, it acts with 

 increased power and velocity. We can continue the most perfectly 

 uniform motion for any length of time by allowing additional fluid 

 to drop very gradually into the copper of the battery; the one cop- 

 per will answer for any number of magnets. If the zinc plates be 

 separate, by insulating these plates with flannel bags we greatly in- 

 crease the power and add to the duration of the galvanic effect. 



Mr. M'Gauley has endeavoured to examine the relative lifting 

 power of magnets of various forms. The following are some of the 

 results : 



The Iron. Three helices of the same wire, each 22 feet in length, 

 were coiled on three different magnets, and the same battery was 

 used. 



lb. oz. 



Magnet No. 1. In length 28 inches, diam. 2 inches, power 6 8 



2. 8 £ 11 11 



T* 3. A magnet having knobs, B and C each 



V y 1-rV inches in length, diameter 1 inch, 



and connected by an arm -H- by -rV> the 

 wire coiled on the knobs, with connect- 

 ing spiral ; power - 2 4 



It was found on another occasion that when the helix did not bear 

 so great a disproportion to the iron, the power of the larger magnet 

 was comparatively very great. To learn the best size for the iron 

 bar suspended in the pendulum of the machine, three forms of keeper 

 were tried. 



No. 1. 54 long, -[$■ thick, 4 wide ; power 4 



2. 5-H- w * 3 8 



3.5* +1- + 7 8 



The large magnet would not lift a steel needle, but lifted a wire of 

 soft iron equal in size to the needle. Perhaps the intensity of the 



