454 report — 1884. 



ness of the thermometers at which the best results will be obtained. This 

 is the place to mention a very interesting and important investigation, 

 lately published by Pi^ofessor W. Foerster, 1 the Director of the Berlin 

 Observatory. He has found that the position of one of the pillars of the 

 observatory is subject to periodical angular displacements, and that these 

 can be represented by means of two periods, an annual one and one of 

 eleven years' duration. This latter has agreed for three sunspot periods 

 with the sunspot curve. Its amplitude is as much as fourteen seconds of 

 arc. Professor Foerster seems to believe that the effect is due to an accu- 

 mulation of heat within the pillar. His explanation seems to me to be 

 contrary to the laws of thei'mo-dynamics, and I think it is much more 

 likely that the change in inclination is due to general disturbance in the 

 level of the surrounding district, which itself would be a consequence of 

 the eleven years period of underground temperature. Comparing together 

 the effects of winter and summer with those of the sunspot period, it is 

 found that a sunspot maximum brings with it a maximum of temperature. 

 The phases show a retardation, as was to be expected. The contradiction 

 of two so well-ascertained effects as those of Koeppen and Foerster is 

 very curious. Its further investigation will no doubt lead to interesting 

 results. 



Professor Balfour Stewart has taken a different line from that of pre- 

 vious workers. He takes as his variable quantity, not the temperature of 

 the place, but the daily range — that is to say, the difference between the 

 maximum and the minimum temperature of the day. He also confines 

 his attention to the shorter periods of sunspot and magnetic inequalities. 

 The duration of these shorter pei'iods has been previously determined. 

 Whether these periods are real or apparent only is not material, as long 

 as we confine our attention to the same period of time. His latest reduc- 

 tions, undertaken jointly with Mr. Lant Carpenter, 2 have led him to the 

 following results : — 



1. Sunspot inequalities around 24 and 2G days, whether apparent or 

 real, seem to have periods very nearly the same as those of terrestrial 

 inequalities as exhibited by the daily temperature ranges at Toronto and 

 at Kew. 



2. While the sunspot and the Kew temperature range inequalities 

 present evidence of a single oscillation, the corresponding Toronto tem- 

 perature range presents evidence of a double oscillation. 



3. Setting the celestial and terrestrial members of each individual 

 inequality so as to start together from the same absolute time, it is found 

 that the solar maximini occurs about eight or nine days after one of tho 

 Toronto maxima, and the Kew temperature range maximum about seven 

 days after the same Toronto maximum. 



Solar Radiation. 



The most direct method to settle the question of variability of the 

 solar radiation would be to measure directly that radiation. The peculiar 

 difficulties which have hitherto stood in the way of continuous records 

 of solar radiation are now gradually being overcome, and we may hope 

 before long to have some decisive evidence either for or against the 

 variability. Our knowledge at present is very vague. Professors Roscoe 



1 Ast. A'ach., No. 2545, p. 1. 2 P/ve. Royal Soc. xxxvii. p. 314. 



