Notices respecting New Books. 867 



million amperes. Other simple cases are considered later in the 

 volume. The assumptions on p. 102 that the current is linear 

 and that the central portion is a straight line, and not the arc of 

 of a circle, are probably concessions to the mathematical difficulties 

 of the case, but the peculiar sense in which the author employs 

 the term "horizontal" should be noticed. Eor instance, in 

 one of the cases on p. 103 we have 7* = 300 and 2-Z=5000 kilo- 

 metres. But the height of the " horizontal '"' portion above the 

 ground will vary in this case from 300 to over 750 kilometres. 

 From statements in the volume, one infers that it is to be 

 supplemented by a subsequent discussion on Earth currents, 

 Auroras and Stormer's mathematical work. Meantime it is 

 difficult to express at opinion on its theoretical results. What 

 now appear to be discontinuities in the reasoning may have a 

 different aspect given to them by the subsequent work indicated. 

 The author has a way of passing from observation to theory, and 

 from theory to experiment, which makes it difficult for the 

 ordinary man to follow the argument or recognize its goal. It 

 has long been recognized that an intimate connexion exists 

 between magnetic storms and auroras when the latter are visible 

 in temperate latitudes. It has also long been known that in the 

 northern hemisphere auroras are most numerous in high latitudes, 

 and that there is much more magnetic disturbance there than 

 further south. The disturbances at Prof. Birkeland's Arctic 

 stations were usually much larger than at the observatories in 

 temperate latitudes, and considering the distance apart of his 

 stations, the differences between the disturbances recorded at 

 them were exceptionally large. The results are interesting and 

 important in themselves, and are obviously favourable to the view 

 that if — as most people suppose — magnetic storms are due to 

 atmospheric electric currents, their principal seat in the northern 

 hemisphere is in the Arctic. But beyond this it does not seem to 

 the writer that anything is proved. It seems to him that what is 

 mainly wanted at the present stage is a more profound examination 

 of the phenomena presented by magnetic storms. Until the laws 

 of distribution of disturbances have been so far ascertained that 

 observation can check theory, a predisposition in favour of a 

 particular theory may be a positive disadvantage to an investigator. 

 Towards the end of the volume there seems a fuller recognition 

 of the fact that the phenomena of magnetic storms are often so 

 complicated that the electric currents capable of producing them 

 must have a distribution whose investigation would require the 

 use of the most advanced mathematics. Considering the extension 

 of view- — or of imagination — which the discovery of radium has 

 produced in physical circles, the size of the currents which Prof. 

 Birkeland postulates and his conclusions on pp. 311-315 as to the 

 expenditure of solar energy will doubtless appear much less 

 startling now than they would a dozen years ago. One cannot, 

 however, but wonder what figures he would have reached if 

 instead of the relatively trifling disturbances of the winter 

 of 1902-3, he had had to deal with those of 1857, or 1870, or 

 even 1892. C. Chree. 



