636 REPORT— 1892. 



The following Reports and Papers were read : — 



1. Bejyort of the Committee on the Meteorological Observations on Ben 



Nevis. — See Reports, p. 68. 



2. Second Eeport of the Committee on Meteorological Photography. 



See Reports, p. '7'^, 



3. Eighth Report of the Committee on Solar Radiation. 

 See Reports, p. 158. 



4. Interim Report of the Committee on Meteoric Dust, 



■5. Twelfth Report of the Committee on the Seismological Phenomena of 



Japan. — See Reports, p. 93. 



6. Nineteenth Report of the Committee on Underground Temperature. 



See Reports, p. 129. 



7. Preliminary Account of Oceanic Circulation, based on the ' Challenger ' 

 Observations. By Dr. A. Bdchan, F.R.8.E. 



8. On the Advantage of making Astronomical Time agree with Civil Time. 

 By Dr. Sandford Fleming, C.M.O. 



9. YorJc Weather from 1841 to 1890. By J. Edmund Clark. 



Fifty years of observations with reliable instruments would appear sufficient to 

 give the annual mean pressure accurately within one-hundredth of an inch ; tem- 

 perature within 1° F. ; rainfall within 2 per cent, of the total fall. Extremes 

 require a much longer period, and monthly means are of course intermediate. 



The means of the three chief values for York are : — Pressure, 29-904 inches ; 

 temperature (9 a.m. and 9 p.m.), 47°-3F. ; rainfall, 25-005 inches. The mean 

 relative humidity has been reduced for only 121 years, for which the value is 85-0. 



FRIDAY, AUGUST 5. 

 The following Papers were read : — 



1. On Leahy Magnetic Circuits. By Dr. H. E. J. Gr. DU Bois. 



In order to obtain some generally reliable result on this question the particular 

 case of a split-ring, magnetised uniformly and peripherically, was investigated 

 theoretically as well as experimentally. The values of the demagnetising factor 

 were found to approximate more and more towards the theoretical values as the 

 magnetising field is increased — as, indeed, they should by a general saturation 

 theorem, first given by Kirchhoff. For a lower range of magnetisation, up to about 

 half the saturation value, the factors are constant, and smaller than the theoretical 

 values. It may be shown that the leakage coefficients, as defined by Hopkinsou, 

 are proportional to the reciprocals of the above factors ; it thence follows that 

 leakage decreases for high magnetisation, contrary to what has hitherto generally 



