TRANSACTIONS OF SECTION A. 457 



tioii, so that we cannot hope greatly to extend the spectrum by takhig elevated 

 observing stations. The limit of the solar spectrum is reached very rapidly, and 

 the spectrum is sharply and completely cut ofi' at about the line U (wave-length 

 2948). From photographs taken at Viesch in the valley of the Rhone and at the 

 Riftelberg, 1910 metres above it, M. Cornu finds the limits to be at wave-lengths 

 2950 and 2930 respectively. 



In the actual absorption of bright line spectra by the earth's atmosphere, 

 M. Cornu observed among others three bright lines of aluminium, which M. Soret 

 calls 30, 31, and 32 (wave-lengths about 1988, 1930, and 18G0), and he found 

 that 32 could not be seen at the distance of 6 metres ; but on using a collimator, 

 and reducing the distance to 1|^ metres, the line 32 became visible, notwithstand- 

 ing the absorption of the extra lens ; at 1 metre, line 32 was brighter than 31 , and 

 at a quarter of a metre 32 was brighter than either 30 or 31. 



With a tube 4 metres in length between the collimator and prism ray 32 is not 

 seen ; but when the tube is exhausted, ray 31 gains in intensity and 32 comes into 

 view, and gradually gets brighter than 31, whilst 30 changes very little during 

 the exhaustion. AVith the same tube he found no appreciable difference between 

 the absorption by air very carefully dried and by moist air, and concludes that this 

 absorption is not due to tlie \'apour of water, and it follows the law of pressure of 

 the atmosphere which shows it to be due to the whole mass or thickness of the air. 

 Also, M. iSoret has shown that water acts very differently ou the two ends of the 

 spectrum, distilled water being perfectly transparent for the most refrangible rays, 

 since a column of water of IIG cm. allowed the ray 2060 in the spectrum 

 of zinc to pass through : on the other hand, water is so opaque to the ultra- 

 red rays that a length of 1 cm. of it reduces the heat spectra of metals 

 to half their length and one quarter of their intensity. 



In concluding my address, I wish to draw attention to some of those magnetic 

 changes which are due to the action of the Sun, and which are probably brought 

 about by means of the ether which conveys to us his radiant heat and light. 



In his discussion of the magnetic effects observed on the earth's surface. General 

 Sabine has shown the existence of diujnal variations due to the magnetic action of 

 the sun ; also the magnetic disturbances, aurora and earth currents, which are now 

 again beginning to be large and frequent, have been set down to disturbances in 

 the sun. 



Although iron, when raised to incandescence, has its power of attracting a 

 magnet very greatly diminished, we have no proof that it has absolutely no mag- 

 netic power left, and with a slight magnetic action the quantity of iron in the suu 

 would be sufficient to account for the diurnal variations of the magnetic needle. 

 During the last few weeks I have been engaged in examining the declination 

 curves for the month of March 1879, which have been kindly lent to the Ivew 

 Committee by the Directors of the Observatories of St. Petersburg, Vienna, Lisbon, 

 Coiuibra, and Stonyhurst. On comparing them with the Kew curves for the 

 the same period, I find the most remarkable coincidences between the curves 

 from these widely distant stations. It was previously known that there was a 

 similarity between disturbances at different stations, and in one or two cases a 

 comparison betv.-eeu Lisbon and Kew had been made many years ago by Sefior 

 Capcllo and Professor Balfour Stewart ; but the actual photogi-aphic magnetic 

 records from several stations have never been previously collected, and so the 

 opportunity for such comparisons had not arisen. Allow me to draw attention to 

 a few of the more prominent features of these comparisons which I ha^e made. 

 On placing the declination curves over one another, I find that in many cases there 

 is absolute agreement between them, so that the rate of change of magnetic 

 disturbances at widely distant stations like Kew, Vienna, and St. Petersburg is 

 precisely the same ; also similar disturbances take place at different stations at the 

 same absolute time. It may be stated generally, for large as well as small disturlv 

 ances, that the east and west deflections of the declination needle take place at the 

 same time and are of the same character at these widely distant stations. 



There are exceptions to this law. Some disturbances occur at one or two stations 

 and are not perceived at another station. Many instances occur where, up to a 



