in opinion as to the direction in the series in which evohition is 

 proceeding — that is to say, whether the red or orange coloured 

 stars are younger, arid become white in old age, or whether the 

 white stars are in the earlier stage, and pass, in the course of ages, 

 into coloured stars. Professor Higgins himself inclines to the latter 

 theory, and to the classification of stars suggested by Vogel in 

 1814, in which the white stars (by far the most numerous) 

 represent the younger and longest stage of stellar life, and the 

 orange and red stars, the period of decline and old age According 

 to some authorities, the sun himself is of about middle age, but 

 it must always be remembered that we do not see him as he really 

 is, his colour being, to our view, materially modified by the inter- 

 vention of the eai'th's atmosphere. iSome experts contend that a 

 star passes successively from the orange and red to the white stage, 

 and then returns by slow degrees to the red and orange. Another 

 theory suggests that the divergence in colour of the stars does not 

 mark any stage of evolution in their existence, but is due to a 

 difference in their original constitution. 



Although the w^ rk of interpreting the nature of the stars or 

 nebulae is comparatively easy so long as their spectra agree with 

 the solar spectrum, it becomes a task of great difficulty when their 

 spectra differ widely from chat of the sun, which, as [ have already 

 shewn, contains so large a number of terrestrial elements. It must 

 also be borne in mind that the lines or waves which the spectro- 

 scope reveals were set in motion many hundreds, or perhaps 

 thousands, of years ago, and the suns whence they proceed may 

 now be dark burnt-out orbs. In the case then of those stars whose 

 spectra differ from that of the sun, there remain but the tests of 

 the laboratory to determine their natm-e, density, and temperature. 

 A spectrum of a remarkable nature is presented by the Aurora 

 Borealis, and although we have reason to believe that certain gases 

 are jiresent in it, no similar spectrum has yet been obtained from 

 them by artificial means. It has recently been suggested that the 

 Aurora was produced by the dust of meteors and falhng stars, which 

 contained traces of manganese, lead, barium, thallium, iron, and 

 other substances. Exhaustive experiments on this point have been 

 made by Dewar and Liveing with the minute dust thrown off from 

 the surfaces of electrodes of various metals, earned forward under 

 the tube of observation by a rapid current of air or gas. It 

 was then shown that metallic dust does not become luminous in an 

 eleccric discharge witli the characteristic spectrum of the Aurora. 



It is now, I believe, generally accepted that the Aurora is a 

 manifestation of the earth's magnetism, and that there is some 

 connection between the display of the northern lights and the 

 maximum period of sun spots. Indeed if the spots or maculsB are 



