200 



REPOKT 1880. 



tennitteutly, the circuit being alternately made and broken by a clock, M. 

 ic, ij arc perfectly similar coils of about 800 ohms' resistance each, and 

 adjusted in position relatively to Z, so that when the condensers on the 

 two sides of the balance have perfectly equal capacity no sound is heard 

 in an especially delicate telephone, T, when the connections are made as 

 in the figure. It will be observed that the nature of the arrangement is 

 such that any failure of insulation in A would make it appear to have too 

 large a capacity and not too small, as would be the case with the method 

 of experimenting previously described. 



For air at pressures gi'eater than one millimetre, the Committee have 

 not thought it necessary to make many experiments, but between 001 

 and 0001 of a millimetre pressure several sets of expeinments have been 

 carried out. At the latter pressure — that is, at about one-millionth of an 

 atmosjDhere — the specific inductive capacity is certainly low, some experi- 

 ments apparently making it as much as 0-6 to 08 per cent, less than that 

 for ordinaiy air, whereas the greatest diminution obtained for an air- 

 vacuum in the two previous investigations, when the pressure was not 

 diminished lower than 5 millimetres, was only O'l per cent. In all the sets 

 of experiments for very low pressures there are decided waves in the curves 

 connecting capacity and pressure, but whether these waves really express 

 a physical law, or whether they are due to the method of experimenting 

 witb currents of very short duration, or whether, lastly, they are due to 

 the capacity not depending solely on the pressure, but also on the amount 

 of residual gas occluded in the aliminium cylinders, the Committee have 

 not yet ascertained, and therefore in this preliminary report they refrain 

 from giving the curves. A sample of the observations, however, may be 



The readings on either day may be compared one with another, but 

 not those on different days. In either case, however, a difference of 100 

 in the reading indicates roughly a difference of about 1 per cent, in the 

 capacity of the aliminium condenser. It will be observed that, in addition 

 to the curious fluctuations in the capacity for the same small pressure, a 



