460 KEroRT— 1882. 



It appears that the longest wave of solar energy which passes through our 

 atmosphere has a length of about 00027mm., or in other words that the spectrum 

 terminates near 27,000 of Angstrom's scale. 



This investigation is supplementary to another, which is designed to give the 

 total absorption of each ray both in the solar and terrestrial atmospheres, for by 

 repeating these measures at great altitudes, and by comparing the absorption at 

 the centre and limb of the sun, for each i-aj-, we can demonstrate the fact that the 

 solar photosphere seen without any absorbing medium would appear deeply blue. 



The method of determining the actual colour of the sun (photosphere) by a 

 special use of Maxwell's colour-discs will be indicated, the approximate result 

 being that this colour is that of the spectrum F. Some conclusions aftectiug our 

 customarily received opinions as to the nature of white light will be indicated. 



7. On a Similarity between Magnetical and Meteor ulogical Weather. 

 By Professor Balfour Stewart, M.A., LL.D., F.B.8. 



It has been hitherto supposed that there is no traceable likeness between the 

 magnetical and meteorological changes of the globe. The former have been 

 imagined to be of a cosmical nature affecting all parts of the earth at the same 

 moment of time, while the latter are well known to be of a local and progressive 

 nature. As a matter of fact all attempts to trace a likeness between simultaneous 

 magnetical and meteorological phenomena have been without success. 



There is, however, one class of magnetical phenomena that are of a progressive 

 nature. I allude to the diurnal variations of the magnetic elements caused by the 

 sun. Of these the solar-diurnal variation of the magnetic declination — that is to 

 say, the variation of the position of a freely suspended magnetic needle, is that 

 which has been most obser\ed and is best understood. 



It has been noticed that the diurnal progress of this variation is not unlike that 

 of atmospheric temperature, the hourly turning-points in both being pretty nearly 

 the same. Both phenomena too are regulated by the local time at the place of 

 observation, and hence are of a progi'essive nature, travelling with the sun in his 

 apparent course from east to west. Both phenomena too are subject to a well- 

 marked annual fluctuation — the diurnal temperature range, for instance, or the 

 difference between the indication of the maximum and the minimum thermometers 

 being greater in summer than in winter, and in like manner the diurnal declination- 

 range or the difference between the east and the west positions of a suspended 

 magnet being greater in summer than in winter. Finally both phenomena appear 

 to be subject to the influence of something which may be called iceather. Some- 

 times we have very hot days and cold dry nights in which the diurnal temperature- 

 range is very groat, succeeded by close rainy weather in which the diurnal tempera- 

 ture-osciUation is very small. In like manner we have sometimes a very large 

 and at other times a comparatively small diurnal oscillation of the magnetic 

 needle, so that it too is affected by the influence of magnetic weather. The 

 question which I now wish to put is the following : Is there any connection 

 between these two weathers ? — between temperature-range weather and between 

 declination-range weather, both defined as above ? Now there is, I think, pre- 

 liminary evidence to show that both kinds of weather are due very greatly, if not 

 altogether, to changes in the sun, a large declination-range and a large tempe- 

 rature-range denoting an increase of solar power. There is also evidence that 

 temperature-range weather once produced travels from west to east, taking 

 probably, on an average, eight or nine days to cross the Atlantic. There is also, I 

 think, preliminary evidence that declination-range weather travels likewise from 

 west to east, but quicker than temperature-range weather, taking about two days 

 to cross the Atlantic. Now if this be true it might be expected that the decli- 

 nation-range weather of to-day should be found similar to the temperature-range 

 •weather six or seven days afterwards, so that by a study of the declination-range 

 weather of to-day we should be able, with a certain measure of success, to predict 

 the temperature-range weather six or seven days afterwards. 



