470 Dr. W» F. G. Swann on the Electrical 



By turning the tap the plate could be slid along the tube, 

 and films could be deposited through the mica slit in various 

 positions on the plate. The discharge was allowed to pass 

 for about 1 hour before the deposition of the first film, and 

 before the deposition of each successive film the vacuum was 

 adjusted to the same value, as indicated by the appearance 

 of a subsidiary discharge tube. In the cases of the longer 

 times of deposit, the deposition was carried on in several 

 stages, the vacuum being adjusted before each stage. After 

 the films were made they were placed in a tube and heated 

 for about 20 hours to a temperature of 115° C. in order to 

 anneal them. 



The resistances of the films of low resistance were measured 

 by a post-office box in the ordinary way. Those of high 

 resistance were measured by tapping off a certain fraction of 

 the e.m.f. of a cell from a potentiometer, and causing it to 

 send a current through the film and through a galvanometer, 

 the apparatus being afterwards standardized by means of a 

 known high resistance. 



The resistances of the films were measured at 100° C. (the 

 tube containing them being in steam), at 14° C. in water, 

 and at —180° C in liquid air. 



It was noticed, especially in the case of the thinnest films, 

 that if a film had been allowed to rest for some time and its 

 resistance was measured at room temperature (14° C. for 

 example), then at 100° C, and afterwards at 14° C, there 

 was a fairly large difference between the two measurements 

 at 14° C. If, however, the film was again raised to 100° 0. 

 and then cooled to 14° C 5 the last two measurements at 

 14° C. agreed very well. In fact, after a film has been 

 raised to 100° C. its behaviour with regard to lower tempe- 

 ratures remains regular for a considerable time. This is 

 well shown by the following set of numbers given in Table I. 

 The table gives the resistances in ohms of the films corre- 

 sponding to the temperatures given in the top line, and to 

 the times of the deposit given in the second column, the first 

 column giving the number of the film for convenience in 

 reference. All these films were of course made in the same 

 vacuum on the same plate. 



In estimating the concordance of these results, no weight 

 must be placed on the results for film (1), since the galvano- 

 meter deflexion corresponding to this film of very high 

 resistance amounted only to about 6 scale-divisions. It will 

 be seen that after the first heating to 99°'8 C. the observa- 

 tions become fairly concordant, and the mean results corre- 

 sponding to the readings taken after the first heating to 



