PRODUCTION OF MECHANICAL EFFECT. 443 



more. The drawing off of the charge for use may 

 be done safely, but somewhat wastefully, in two 

 hours, and very economically in any time of from 

 five hours to a week or more. Calms do not last 

 often longer than three or four clays at a time. 

 Suppose, then, that a five clays' storage-capacity 

 suffices (there may be a little steam-engine ready 

 to set to work at any time after a four-days' calm, 

 or the user of the light may have a few candles or 

 oil-lamps in reserve, and be satisfied with them 

 when the wind fails for more than five days). One 

 of the twenty kilogramme cells charged when the 

 windmill works for five or six hours at any time, and 

 left with its sixty-candle hours' capacity to be used 

 six hours a day for five days, gives a two-candle 

 light. Thus thirty-two such accumulator cells 

 so used would give as much light as four burners of 

 London sixteen-candle gas. The probable cost of 

 dynamo and accumulator docs not seem fatal to 

 the plan, if the windmill could be had for some- 

 thing comparable with the prime cost of a steam- 

 engine capable of working at the same horse-power 

 as the windmill when in good action. But wind- 



