424 



NATURE 



[August 28, 1890 



It was then found that r6 volt was far too low a limit 

 to take, as scaling of the plates took place, and so (as 

 mentioned above) limits of i"8 and 2-4 volts were adopted. 

 After several charges and discharges, the cells arrived at 

 a "steady state" again, the successive times being as 

 follows : — 



Charges, 

 h. m. 



II 38 



II 37 



II 37 



" showing to what an absolutely definite state cells arrive 

 after a definite cycle of charge and discharge between 

 fixed limits has been repeated continuously, without in- 

 terruption, for some weeks." These results give an 

 ampere efficiency of 97-2 per cent, and an energy 

 efficiency of 87 "4. These they adopt as the true steady 

 values for this type of cell, and this shows a working 

 storage capacity of 21,380 foot-pounds per pound of 

 plate. 



The next point brought out is the effect of rest on a 

 charged cell ; the cells were fully charged, and allowed 

 to rest in that state, being well insulated from everything. 

 In every case the first discharge and charge show a 

 marked falling off in capacity and efficiency, the latter 

 being reduced to 58 per cent, in one case cited. A point 

 of some theoretical interest is brought out in connection 

 with the curves obtained in this part of the investigation : 

 a normal discharge curve falls rapidly at first, then re- 

 mains constant, and falls again, as shown in Fig. i ; 



the first discharge curve, a/ler rest, rises at first instead of 

 falling. 



The authors sum up this part of their paper as 

 follows : — 



" From all that precedes, it follows that the previous 

 history of an accumulator produces an enormous effect 

 on its efficiency. If, for example, an E.P.S. accumulator 

 be over and over again carried round the cycle of being 

 charged up to 2*4 volts per cell and discharged down to 

 1-8 volt per cell, the charging and discharging currents 

 being the maximum allowed by the makers — viz. 0026 

 ampere per square inch in charging, and 0-029 ampere 

 per square inch in discharging — the *■ working efficiency'' 

 thus obtained may be 97 per cent, for the ampere-hours 

 and 87 per cent, for the watt-hours. If, on the contrary, 

 the cell be constantly charged up before being tested, 

 then for the first few charges and discharges between 

 the above limits, and with the same current-density 

 in charging and discharging, even the energy efficiency 

 may be as high as 93 per cent. ; whereas, if the accu- 

 mulator has been left for some weeks, then, although 

 it was left charged, the energy efficiency for the first 



NO. 1087, VOL. 42] 



few charges and discharges will be as low as 70 per 

 cent. 



" While, on the one hand, our tests show that continued 

 rests of a charged accumulator appear to be far more 

 serious for the accumulator than we had previously 

 imagined, the working efficiency appears to be higher 

 than has hitherto been supposed, since we believe that 

 about 84 per cent, efficiency in the watt-hours is all 

 that the advocates of accumulators have claimed for 

 them." 



The next section deals with some points connected with 

 the chemical action, and it is shown that the actual amount 

 of SO3 liberated on charge per ampere-hour, as calculated 

 from the change of specific gravity, agrees well with the 

 ordinary simple formulae. We understand a further paper 

 on this point may probably be forthcoming later on, which 

 will deal with the chemical changes going on in the 

 plugs at various points in the charge and discharge. As 

 this involves the partial destruction of a cell, and a lengthy 

 series of analyses, it was not found possible to put it in the 

 present paper. 



The next point brought out is of considerable interest : 

 during several charges and discharges, the difference of 

 temperature between the working cell and a neighbour- 

 ing idle cell was observed frequently, and it was noticed 

 that the cell cooled during discharge, in spite of the heat 

 generated by the resistance. This was simultaneously 

 observed by Prof. Duncan in America. The general 

 shape of the temperature curves is given below. 



^C Varlalion or Temperature during Discharge ana Charge 



7. 



Z 



6' 



z 



/ 



2 3 4 5 G 7 8 



Tmt in Hours from Beginning of Discharge or Charge 



From the mean excess of temperature of the cell 

 the authors deduce the somewhat startling fact that 

 17 per cent, of the energy put into the cell is wasted 

 by radiation and convection. As they found that but 

 13 per cent, is really lost, it follows that the rest of 

 the energy must be given by some sort of primary 

 battery action, so that they consider an accumulator 

 is partly a reversible and partly an irreversible battery. 

 In this way the gradual deterioration is accounted for. 

 Possibly this may partly account for the short life of 

 small accumulators. 



The concluding section deals with the question of the 

 resistance of cells when brought to a steady state. The 

 method adopted is that of introducing an opposing 

 E.M.F. in the voltmeter circuit when time readings of 

 the E.M.F. are being taken on breaking circuit. From 

 these the E.M.F. at breaking circuit is found by pro- 

 ducing backwards to zero, and the P.D. being also 

 measured, together with the current before breaking, we 



