]90 



KNOWLEDGE. 



[October 2, 1893. 



in 1878 found ihe surface of the sun to be five thousand 

 three hundred times brighter than the molten metal in a 

 Bessemer " converter." There is evidently a great range 

 in the light-giving capacity of incandescent bodies, but the 

 briUiance of the solar surface does not greatly exceed the 

 brightest incaudescence of the most refractory bodies, ob- 

 tainable by means of the electric current ; and it should be 

 remembered that in laboratory experiments the intensely 

 heated part of a glowing rod of carbon, or other refractory 

 substance, is being cooled by adjacent cold parts, and it is 

 therelore probably not as hot as if the experiment were 

 made on the solar scale. 



Such considerations would lead us to conclude that the 

 light of the photosphere must be due to the brilliant 

 incandescence of the most refractory substances present 

 in the sun, at a level w^hen they are just on the point of 

 being driven into vapour. 



To those who would urge that we do not know what 

 would be the brightness of an explosive flame two thousand 

 miles thick. I would reply that such explosions could not 

 go on continuously over a wide area — there must be room 

 for the descent of the associated products of combustion, 

 and an interval of time for the ascent and cooling to the 

 temperature of explosive combination of the ascending 

 elements ; also, that in such a mixed mass of vapours as 

 must exist in the well-chm-ned region of the solar photo- 

 sphere, the brightness of the explosive flame would pro- 

 bably be considerably less than the brightness of the flame 

 where pure gases combine; and further, that a flame 

 containing metallic vapours, such as we know are present 

 in the region immediately above the photosphere, would 

 give a spectrum showing many bright lines, which it 

 S( ems probable would not all be simflarly reduced by 

 absorption, so as to give the appearance of a comparatively 

 uniform continuous spectrum, crossed by dark lines, such 

 as we see when the spectrum of the light of the photo- 

 sphere is examined. 



It has recently been suggested by more than one phy- 

 sicist that it is possible that gases only give out their 

 characteristic bright-line spectra when some chemical or 

 electrical change is going on in their molecules ; but it 

 seems to me that this suggestion (and it is nothing more 

 than a speculative suggestion) is suiticiently negatived by 

 facts already in our possession. Prof. Smithells has him- 

 self >hown that iodine vapour may be rendered incandescent 

 by merely beating it in a glass tube, warmed to a red glow 

 in a Bunsen burner ; and following the analogy offered by 

 the radiation of solid bodies, we should expect that a 

 gas which absorbs as freely as iodine vapour would also 

 ladiate freely, and that other vapours which absorb 

 radiations less greedily than iodine vapour would only 

 commence to radiate when raised to temperatures higher 

 than can be conveniently obtained for laboratory experi- 

 ments. 



The phenomena of the sun itself also seem to indicate 

 that the incandescence of vapours in the chromosphere is 

 not due to chemical changes, but is more probably a mere 

 heat eflect. If the incandescence were due to chemical 

 change, we should expect all the bright lines corresponding 

 to any particular element to extend to the same height 

 above the photosphere, namely, to the height at which the 

 chemical change affecting the element was taking place ; 

 but the lines in the spectra of the same element extend to 

 very dift'erent heights, and that not always in the order of 

 their brightness. 



Hitherto solar physicists have generally atttiiipted to 

 account for the spectral phenomena presented by the sun 

 as mere heat effects caused by radiation from hot matter, 

 and absorption by cooler gases ; and it is no doubt proHt- 



able to pursue this method without making assumptions 

 as to unknown electrical efl'ects, and the chemistry of 

 temperatures with which we are not familiar in terrestrial 

 laboratories. But it must be remembered that more than 

 half of the lines of the known solar spectrum remain 

 imaccounted for, and that no lines corresponding to whole 

 groups of chemical elements have as yet been recognized 

 in the solar spectrum. 



In the last number of Knowledge I endeavoured to show 

 that the vapoiu's present in the chromosphere and corona 

 cannot form a solar atmosphere in which gases rest in 

 equDibrium stratified into absorbing layers, as is so 

 frequently assumed. The proof turns upon what we know 

 of the intensity of solar gravity at the level of the photo- 

 sphere, and upon the assumptions which it seems safe to 

 make with regard to the temperature of the region 

 immediately above the photosphere. If the photosphere 

 consists of incandescent clouds of liquid or solid particles 

 of matter similar to the matter of which the earth is 

 composed, the general temperature of the region cannot be 

 much above the temperature at which similar matter would 

 be driven into vapour in terrestrial laboratories. It may 

 be a little above such a temperature, because the matler 

 of which the photospheric clouds are composed may be 

 continually falling from a cooler outer region towards the 

 hot solar nucleus. 



Two classes of observations made during total solar 

 eclipses tend to support the theory that the photosphere 

 is composed of incandescent solid or liquid particles, for 

 they show that it is highly probable that minute particles, 

 many of which have diameters that are small compared 

 with the wave-length of light, exist in the corona and 

 extend down into its lower regions. When the corona is 

 examined with a suitable polariscope, strong polarization 

 colours, which extend down to close to the moon's limb, 

 are observed, indicating a condition of radial polarization 

 of the coronal light such as can be accounted for by dust 

 polarization in the region of the corona. Secondly, the 

 solar absorption lines have been observed in the coronal 

 light down to the brilliant region near to the moon's limb, 

 indicating that the corona chiefly shines with light which 

 has been received from the sun and has been dispersed by 

 liquid or solid matter in the coronal region. Most 

 chemists and physicists will therefore, no doubt, agree that 

 we are reasonably safe in assuming that the temperature 

 of the coronal region cannot exceed 10,000° Cent. 



The narrow character of the lines in the spectrum of the 

 chromosphere seems to indicate that the gaseous pressure 

 in the region where they exist must be very small com- 

 pared with the pressure of the earth's atmosphere at the 

 sea-level ; but assuming that the gaseous pressure at the 

 level of the photosphere is e(jual to the pressure of our 

 atmosphere at the sea-level, and assuming the height of 

 the solar atmosphere to be the height at which the pressure 

 is reduced to a millionth of a millionth of the pressure at 

 the level of the photosphere, we can calculate the height 

 of a hydrogen atmosphere at any temperature with solar 

 gravity equal to 27-f times terrestrial gravity. Thus, at a 

 temperature of 2457° Cent., the height of a hydrogen 

 atmosphere about the photosphere would be only 270 

 miles, and at a temperature of 27,027° Cent, its height 

 would be 2700 miles. 



It will then, no doubt, be asked, can we, with such a 

 rapid decrease of densities as is indicated by the above 

 results, retain the old idea that the photosphere is a layer 

 of clouds floating in the solar atmosphere ? We know 

 nothing of the temperatures within the photosphere, and 

 without a knowledge of such temperatures we cannot 

 determine the depth at which the density of the solar 



