258 ANNUAL KEPORT SMITHSONIAN INSTITUTION, 193 3 



electrolysis by the use of gravitational energy alone, and set up a 

 device somewhat similar to the Kelvin water-dropper. My device 

 consisted of drops of copper sulphate coming from an outlet like 

 that in the cylinder A and falling into a platinum funnel like that 

 directly below A. I charged the cylinder A with a large negative 

 charge from a static machine, and this charge remained on the 

 cylinder, which was well insulated. Consequently, every drop of 

 copper sulphate that dropped carried an induced positive charge and 

 delivered it to the funnel 6, which was earthed. This excess positive 

 charge would of course be in the form of copper ions which would 

 be deposited on the platinum in the process of neutralization of the 

 drop. After running the apparatus for an hour or so, I looked at the 

 platinum cylinder to see whether I could see any copper deposited 

 on its inside, and finding none, I set the apparatus going in the late 

 afternoon and let it run automatically until the following morning. 

 Again examining the funnel, I found no deposit of copper and, some- 

 what surprised, I sat down to figure. I soon discovered that the 

 copper would be present in far too small a quantity to detect. In 

 fact, if every drop were charged with the largest amount of elec- 

 tricity which it could carry without losing it by corona to the sur- 

 rounding air, and if the drops had fallen as fast as possible begin- 

 ning with the time of Christ, I would by this time have collected 

 barely enough copper to be shown by the most sensitive known 

 chemical test. This little experience illustrates the vast difference 

 in magnitude between the kind of currents that we are accus- 

 tomed to deal with in electromagnetic induction devices, dynamos and 

 motoi"s, and these relatively very feeble currents of electrostatics. 

 These drops were charged with high electrostatic voltage, and the 

 device w^as a fairly efficient electrostatic generator, and yet 2,000 

 years would have been required to deposit an amount of copper such 

 as would appear in a fraction of a second with only a moderate 

 current of the type that we ordinarily use iu electromagnetic 

 instruments. 



In recent j^ears an interesting development of the Kelvin water- 

 dropper has been proposed by Dr. Swann, of the Bartol Research 

 Laboratory, in which the water drops are replaced by steel balls 

 which fall under the action of gravity, and in order to make the 

 process continuous, there is the suggestion that these balls may be 

 carried back again to the upper container by means of magnetic 

 control. In this way the succession of falling balls behaves some- 

 what like a continuous belt containing metal sections separated by 

 insulated regions of air and driven by gravity. In the absence of 

 leakage this kind of a generator should be capable of developing such 

 a high voltage that the electrostatic attraction of the fallinjr balls 



