452 report — 1884. 



primarily only by its radiation. Tins is the idea which forms the founda- 

 tion of Professor Balfour Stewart's work on the subject. He has also 

 thrown out the idea that connection currents in the outer part8 of our 

 atmosphere may, by their motion across the lines of magnetic force, induce 

 sensible currents. I do not believe that our present knowledge of the 

 passage of electricity through gases quite warrants the assumption of 

 such a possibility, for though we may make the electric resistance in 

 some parts of the vacuum almost as small as we please, the total 

 electro-motive force which is capable of setting up a discharge at all must 

 always be considerable. At any rate, it is absolutely necessary, before 

 going further, to decide whether we can trace a connection between the 

 number of sunspots and those terrestrial phenomena which depend ou his 

 temperature. It is only when we are in possession of all the facts that 

 we may hope for a solution of the mystery. 



Connection of Stinkpots and Temperature. 



We owe to Gautier ' the first detailed examination of this question, 

 which led him to the result that the years of many sunspots are probably 

 rather colder, those with few sunspots rather warmer, than the average. 

 Other investigations led generally to a similar, occasionally, however, to a 

 contradictory result, until Koeppen, by means of a very exhaustive in- 

 vestigation, has given us some decisive results. Koeppen 2 has brought 

 together the temperature records of nearly 250 stations from different 

 parts of the earth. 



These were divided into five groups, according to their geographical 

 position — namely, tropics, sub-tropics, the warmer parts of the temperate 

 zone, the colder parts of the temperate zone, and the cold zone. Each of 

 these groups was examined separately. The curves are plotted down in 

 Plate II., and will show a very remarkable relationship to the sunspot 

 curve. The connection is most marked for the tropics, then gradually as 

 we move away from the equator it becomes less and less distinct. The 

 following table, which compares together the years in which the maxima 

 and minima took place, will also render the connection very clear. The 

 table in which the ectropical regions have been united is taken out of 

 Hahn's monograph. It is well known that sunspot periods differ in length, 

 and that the time between a maximum and a minimum is shorter than 

 that between a minimum and a maximum. This peculiarity is exactly 

 reproduced in the temperature curve, as the comparison (Table II.) made 

 by Koeppen will show. The numbers express the years intervening 

 between two successive turning-points of the curve. 



The difference between the average temperature in a year of maximum 

 and one of minimum sunspots is by no means small; it reaches o- 73 C. 

 in the tropical regions, and a little over half a degree in the ectropical 

 groups. The maxima and minima of temperature in the tropics seem to 

 take place a little before the corresponding phases of the sunspot curve, 

 while the ectropical regions show a retardation of the phase. The curve 

 shows some irregularities about the year 1860, bat especially towards the 

 beginning of the century it is very much disturbed ; and, worse still, 

 between 1780 and 1790 the effect seems exactly reversed. This latter 



' Ann. Clam. Phys., III., XII. (1844). 



2 Zeitschrift der oext. Ges. Met., vol. viii. pp. 241 and 257. 



