380 SECTIONAL TRANSACTIONS.— G. 



Many undertakers will be able to purchase electricity in bulk at a considerably 

 less cost than that at which they could generate for themselves, and will, therefore, 

 •concentrate their whole energies on the selling of electricity, which, after all, is their 

 proper and legitimate sphere of activity. 



Electricity in bulk (probably in the form of 3-phase alternating current at 

 11,000 volts and 50 periods) then becomes the raw material of these distributing 

 agencies, whose function it will be to transform it into the form of energy required 

 by their consumers and to distribute it throughout their area of supply. 



The success or otherwise of a distribution undertaking may be judged by two 

 main principles, as follows : — 



(1) The difference between the average revenue and the average purchase-price 

 per unit sold should be a minimum. 



(2) The number of units sold per £ of capital invested should be a maximum for 

 ■each class of distribution undertaking. 



The paper deals with the problems involved in following these principles, which 

 not only include the technical questions relating to transformation and distribution, 

 tut also the consideration of the most desirable forms of tariff for the bulk supply 

 and for the different classes of consumers in any district. 



An analysis is given of the various costs incidental to electrical distribution. The 

 ■effect on load factor and revenue of a properly balanced rate system is pointed out, 

 together with the cumulative increase in number of units sold per head of population 

 resulting from each decrease in the price charged for the various classes of supply. 



5b. Mr. BoRLASE Matthews. — Electric Plouc/hing. 



6. Prof. Wm. Cramp. — Some Phenomena of Electric Conduction. 



An electric current is generally held to be due to the relative motion of protons 

 and electrons. Thus if — e be the electronic charge, N the number of protons or of 

 ■electrons per cubic cm., U the drift of the protons, and V the parallel drift of the 

 electrons, then the current density is 



-NeV+NeU = Ne (U-V). 



This argument receives support from the work of Rowland, but his experiments 

 •are not conclusive. It is also in accord with the general facts of electrolysis. But 

 there is no experimental proof of the assumption that a negative charge of given 

 magnitude moving at a given speed produces the same magnetic effect as a similar 

 positive charge moving at the same speed in the opposite direction. In metallic 

 •conduction it is generally supposed that the electrons only drift. In electrolytic 

 conduction both anions and cations drift. The question arises, therefore, ' Is the 

 magnetic field produced by an electrolytic conductor the same as that obtained 

 when a metallic conductor of similar form carries the same current ? ' Sheldon and 

 Downing in 1898 attempted experiments to settle this question, and found no 

 difference between the two cases ; but their method is open to much criticism. 

 Physics text-books generally state that there is no difference, but they give no 

 proofs. The author's plan was to compare the magnetic field produced by a tube of 

 ■electrolyte carrying a given current with that produced by the same current when 

 the tube was full of mercury. The electrolytes used were respectively silver nitrate 

 and hydrochloric acid. Two methods of measurement were adopted for the 

 magnetic field, viz. (1) the ballistic -galvanometer method, and (2) the direct-reading 

 magnetometer. In all cases the magnetic field due to the silver nitrate was less than 

 that due to the mercury, and that due to the hydrochloric acid was slightly less than 

 that due to silver nitrate. Further, bj' both methods the percentage difference was 

 very nearly the same, viz. just under 2 per cent, as between mercury and hydro- 

 chloric acid. 



The possible causes of this difference have been examined. It is found that if 

 the field were produced by electrons only the differences should be far greater. If it 

 were due to the fact that in the electrolyte the drifting charges are loaded with 

 drifting masses of matter, while in the mercury they are not, the differences should be 

 much less. The results are not in harmony with the supposition that there is a 

 difference between the magnetic field produced by a proton drift and that due to an 

 electron drift. The author was finally driven to the conclusion that the difference is 

 due to the fact that, in the metal, conduction is due to an electron drift of comparatively 



