964 Mr. Beer and Prof. Tyndall on Manometric 



the true pressure at the surface of the electrode will only be 

 recorded by using a hole which is small compared with the 

 thickness of this layer. 



The evidence given above is definitely against (1). Our 

 experiments furnish no evidence for or against (2), whereas 

 it can be shown that they give strong quantitative support 

 for (3). 



The experimental troubles brought about by the constric- 

 tion due to a fine hole place a lower limit on the diameter 

 which can be successfully used, and, except in a few cases 

 (when the pressure rose still higher), no hole of diameter 

 less than 1*7 mm. has been experimented with — a length 

 which is almost certainly enormously greater than the 

 thickness of the pressure layer. 



Given, however, that the pressure-producing layer is very 

 thin and that the current density at the luminous patch is 

 independent of the size of the hole, it is easy to see that the 

 pressure observed at a relatively large hole at the centre of 

 the patch is approximately proportional to the circumference 

 of the hole at which the pressure is communicated and 

 inversely proportional to the area of the hole. In other 

 words, we might expect on hypothesis (3) that the observed 

 pressure will be approximately inversely proportional to the 

 diameter of the hole. When the hole is comparable in size 

 with the luminous patch, it is impossible to obtain true cen- 

 trality of hole, but it is reasonable to extend the argument 

 to the case of a displaced hole, provided the conditions are 

 similar for all the holes used. An analysis of the following- 

 results for different currents and sizes of hole substantially 

 verify the hypothesis. 



In every case, whatever the size of the hole, an increase 

 of current is accompanied by an increase of pressure at the 

 anode. Typical results are shown in curves IV. The values 

 of size of hole and length of arc applying to curves A, B, 

 and C are given on the graph. 



Graph D is the only line passing through the origin. 

 The readings for this were taken under conditions in which 

 the greatest care was observed in ensuring centrality of the 

 arc, the hole being always smaller than the luminous patch 

 so that this was possible. In curves A, B, and 0, on the 

 other hand (except perhaps at the largest currents), the hole 

 was always larger than the luminous patch and the centreing 

 consequently not perfect. 



The conditions of centreing were, however, sufficiently 

 similar to enable us to use the three curves, A, B, and 

 (as well as others which for sake of clearness are not 



