322 BEPOET — 1887. 



(2. ) Temperattore-coefficients. 



In the opinion of this Committee the best method of determining the 

 temperature-coefficients of the two force magnetographs is by alternately 

 heating and cooling the magnetograph room, the former operation being 

 conducted by means of a stove devoid of iron. Two sets of experiments 

 are desirable, one in summer and one in winter. The Committee are 

 likewise of opinion that it is necessary to obtain separate coefficients for 

 ascending and descending temperatures, inasmuch as the behaviour of 

 magnets with respect to temperature is not perfectly reversible. 



More especially will this be necessary in the case of the Balance, or 

 vertical force magnet, for if this has its temperature-coefficient partly 

 corrected by means of a zinc bar the uncompensated portion may be 

 very different for ascending and descending temperatures. The director of 

 any observatory who is willing to co-operate with the Committee is requested 

 to communicate with Professor W. G. Adams, King's College, London. 



The Rev. S. J. Perry and Professor Stewart are continuing their com- 

 parison of simultaneous magnetic fluctuations at Kew and Stonyhurst. 



Professor Schuster has recently communicated to the Royal Society 

 a paper entitled ' Experiments on the Discharge of Electricity through 

 Grases,' which bears upon the work of this Committee. 



These experiments show ' that a steady current of electricity can be 

 obtained in air from electrodes at the ordinary temperature, which are at 

 a difference of potential of one-quarter of a volt only (and probably less), 

 provided that an independent current is maintained in the same closed 

 vessel.' 



Professor Schuster makes likewise the following remark : ' I have last 

 year obtained, by calculation, results which seem to show that the prin- 

 cipal cause of the diurnal variation of terrestrial magnetism is to be 

 looked for in the upper regions of the atmosphere. Professor Balfour 

 Stewart at various times suggested that the aif-cnrrents in these regions 

 may, owing to the lines of force of terrestrial magnetism, have electric 

 currents circulating in them. The difficulty against this supposition 

 always seemed to me to lie in the fact that the electromotive forces 

 required to start a current were larger than those which could possibly 

 exist in the atmosphere. But as there are very likely continuous electric 

 disturbances going on, such as we observe in auroras and thunderstorms, 

 the regions within which these changes take place would act as con- 

 ductors for any additional electromotive force, however small, so that any 

 regular motion, such as tidal motions, could very well produce periodic 

 effects affecting our magnetic needles. 



' If these original discharges increase in importance, then, according to 

 the results obtained in this paper, the currents due to the smaller periodic 

 causes would increase also, and they may increase in a very rapid ratio. 

 We know that the electric discharges in the upper regions of the atmo- 

 sphere are considerably stronger at times of many sun-spots ; and this 

 may account for the fact that at those times the amplitude of the daily 

 oscillation of the magnetic needle is considerably increased.' 



There are six appendices attached to this report. The first and second 

 of these are by Dr. Buys-Ballot, describing his method of separating be- 

 tween disturbed and undisturbed magnetic observations. The third is a 

 list of stations at which magnetic observations liave been made, compiled 

 by Sir J. Henry Lefroy and Mr. "Whipple. The fourth is a continuation 



