384 



NA TURE 



[August 18, 1892 



South Atlantic Oceans at that depth. At 200 fathoms the tern- 

 perature of the Mediterranean is about 56°, and is practically 

 ■constant down to the bottom {1500 fathoms in some places). 

 Similar conditions hold in the Gulf of Mexico, where the tem- 

 perature at 700 fathoms is 25°'5, with no change at lower 

 depths. On the other hand, north of the Wyville Thomson ridge 

 in the North Atlantic, there is a uniform temperature of about 

 29° '5 at all depths below 700 fathoms — that temperature being 

 about two or three degrees higher than the freezing point of the 

 water. This undercurrent of warm salt water from the Medi- 

 terranean, extending even beyond the North Cape of Norway, 

 seems to explain why there is no instance of an iceberg appear- 

 ing off the west coast of Europe. 



Physical Condition of the Waters of the English Channel, by 

 Mr. H. V. Dickson. — The constitution of the samples of water 

 agreed, on the whole, with that of the Challenger samples, 

 coinciding entirely with that of the Atlantic water. The tidal 

 currents are sufficiently strong to keep the water thoroughly 

 mixed from the top to the bottom, except off Start Bay, where a 

 vortex is formed — the water being colder in summer, and 

 warmer in winter, than the surrounding water, and this spot is 

 one of the best fishing grounds in the Channel. 



On Primary and Secondary Cells in which the Electrolyte is 

 a Gas, by Prof. Schuster. — When an electric discharge passes 

 through a part of the gas filling a tube, all the gas is brought 

 into a state in which it readily conducts electricity. Prof. 

 Schuster has studied the laws of this conduction. If we assume 

 that the primary phenomena of discharge depend on dissoci- 

 ation of the molecules, he remarked that it must often have 

 appeared peculiar to experimenters that no phenomena of 

 polarization appear. When an elementary gas is used no such 

 phenomena appear. When compound gases are used only 

 slight polarization appears ; but Prof. Schuster has found that 

 the phenomena becime very marked when hydrocarbons are 

 used. The law of decrease of polarization in this case resembles 

 that which is observed when liquid electrolytes are used. This 

 points to the performance of work of the nature of electrolysis. 

 The magnitude of the effect depends on the nature of the 

 electrodes. It is small when copper and iron are used ; but is 

 very large when aluminium or magnesium are used. When a 

 latter metal was employed, a direct current being passed for a 

 long time, a reverse electromotive force of 35 volts was 

 got from a single cell. This shows that the action is similar to 

 that of a secondary cell. Prof. Reinold has already described 

 •cases in which gases act as an electrolyte in a primary circuit 

 when under the influence of a discharge. Prof. Schuster has 

 found that, in such cases, the employment of aluminium 

 electrodes gives very strong effects. 



On Leaky Magnetic Circuits, by Dr. du Bois. — It appears 

 from the experiments described that the leakage decreases when 

 the magnetization is increased. 



Experiments on the Electric Resistance of Metallic Powders, by 

 Dr. Dawson Turner. — It is well known that metallic powders 

 have very great electric resistance. This can be reduced to an 

 extraordinary degree by the passage of an electric spark in their 

 neighbourhood. Amongst other substances Dr. Turner has 

 tried powdered aluminium, copper, annealed selenium, iron 

 filings, small shot, mixtures of aluminium and resin fused into 

 a solid mass, etc. The best results were obtained with the first 

 two. A short glass tube, filled with powdered aluminium, is 

 placed in circuit with one or two cells and a galvanometer. 

 No current passes until a spark discharge occurs in the neigh- 

 bourhood, when a fairly large effect becomes visible. The 

 powder continues to conduct for a short time unless it be shaken 

 or disturbed, when the effect ceases. In the case of the rod of 

 aluminium and resin, mere shaking does not destroy the effect, 

 though the application of heat does. When the resistance has 

 once been lowered in this way, the powder becomes very sensi- 

 tive, a spark at a great distance produces the effect, and a very 

 slight jar destroys it. 



On the Stability of Periodic Motions, by Lord Kelvin. — The 

 mathematical investigation of this subject was illustrated by 

 an experiment in which a simple harmonic vertical motion 

 was given to the point of support of a pendulum. When the period 

 of the superposed motion was one half of that of the natural 

 motion of the pendulum, the equilibrium became unstable, and 

 the slightest disturbance caused the vertical motion of the bob 

 to be changed into transverse motion of increasing amplitude. 

 If the superposed period were now lessened, the vertical motion 

 again became stable. Similarly a rod poised vertically in un- 



NO. 1 1 90, VOL. 46] 



stable equilibrium could become stable by having its point of 

 support moved with simple harmonic motion, of proper period, 

 in a vertical line. 



Prof. Osborne Reynolds remarked that it was well known to 

 practical engineers that a revolving shaft, when driven at a 

 certain speed, began to bend, and might even break, though at 

 higher speeds it would again become straight. Lord Kelvin had 

 now explained this effect. 



On the Specific Conductivity of Thin Films, by Profs. 

 Reinold and RUcker. — When the film was an aqueous soap 

 solution containing a considerable portion of glycerine and a 

 small proportion of a metallic salt, the specific conductivity was 

 the same, whether the liquid was in mass or was drawn out into 

 a film not exceeding 1-200,000 in. in thickness. When the 

 liquid consisted of an aqueous soap solution alone, the specific 

 conductivity increased when the thickness became small, until, 

 in the thinnest film observed, it was seven times as great as at 

 first. The effect seemed to be due to a breaking down of 

 equilibrium when the tenuity was extreme. 



A Contribution to the Theory of the perfect Influence Machines, 

 by J. Gray, B. Sc. — The theory of the perfect influence machine 

 has been shown by Clerk Maxwell to be analogous to Carnot's 

 theory of the perfect heat-engine. Maxwell points out that there 

 is a loss of energy in the ordinary influence machine through 

 sparking at the contacts, which would render the machine in- 

 efficient, even though losses from leakage and the like were done 

 away with. Maxwell has described a machine in which spark- 

 ing and the loss due to it is eliminated. This is done by causing 

 the carrier of electricity always to make contact with charging 

 and discharging conductors when the former is at the same 

 potential as the latter. In the case of the discharging con- 

 ductor, this is done by prolonging the contact springs to meet 

 the carrier ; in the case of the charging conductors this is not 

 sufficient ; it is necessary to surround the ends of their contact 

 springs by two additional conductors charged to an equal and 

 opposite potential, and of such capacity as will just reduce to 

 zero the potential of the small quantity of electricity left from 

 the previous discharge. These additional conductors were called 

 by Maxwell regenerators, as being analogous to regenerators in 

 the heat-engine. 



The object of the author is to investigate the efficiency of an 

 influence machine constructed according to Maxwell's design, 

 and other designs less perfect. This is done by drawing a QV 

 (quantity-potential) energy diagram for one revolution of the 

 carrier. The results obtained are as follows : — 



Theoretical . 

 efficiency. 



Maxwell's machine i 



without regenerators ... 



without long contacts on re- 

 ceivers 



without regenerators and with- 

 out long contacts on chargers 



Q + ioz* 



Q 



where Q = quantity received or discharged by a carrier in each 

 revolution, 



V = the potential (numerical value) of the positive or 



negative receiver, 



V = the potential of the residual charge to be reduced to 



zero by the regenerator, 

 o = the part of the carrier's capacity due to its not 

 being completely surrounded by the discharging 

 conductor. 

 The conclusion is that the regenerators are of much less im- 

 portance than the long contacts in adding to the efficiency of an 

 influence machine. 



Experiments with a Ruhmkorff Coil, by Messrs. Magnus 

 Maclean and A. Gait. — The quantity of electricity induced in a 

 secondary circuit by a make in a primary circuit is equal to the 

 quantity induced in the same secondary by a break in the pri- 

 mary. If, however, there is a non-metallic gap in the circuit, 

 the break impulse causes a flow in one direction, and the make 

 causes either no flow, or a much less flow in the opposite direc- 

 tion ; because the short intense impulse of the former breaks 

 down the resistance, while the comparatively long and less in- 

 tense impulse of the make either does not break down the re- 

 sistance at all or only does so to a slight extent, so that the 

 effective resistance is much greater in one direction than in the 



