358 



KNOWLEDGE ♦ 



[October 1, 1886. 



as he suggests, nt least in lecognisable degi-ee. In fact, their 

 appearance and changes of appearance suffice entirely to 

 negative this view. 



I have long seen reason to believe that the explanation of 

 the peculiarities of appeaiance observed when a spot is near 

 the sun's edge (or " limb," as astronomers call it, comparing 

 it not very reasonably with the limb of a sextant or quadrant) 

 is to be found in the refractive action of the vapours occupy- 

 ing the depression which the sun-spot really forms beneath 

 the level of the photosphere. 



In the following figures (from my " Sun ") I illustrate 

 what is observed, as well as what ought to be observed, 

 under different assumptions ; — Fig. 1 shows in the upper 



Fig, 



1.— Spots on the Su.v— Actual axd Theoretical 

 Chakges. 



row the changes recognised by Dr. Wilson ; in the lower 

 row the features which would be seen if the spots were 

 Buiface stains with umbia and penumbra symmetrical. 



FiG. 2.— Wilsons Explanation of Sun-Spot Changes. 



Fig. 2 illu.strates the explanation of Wilson's observations 

 commonly advanced, and first suggested by himself. Fig. 3 

 illustrates the feasibility, so far as Wilson's observations are 

 concerned, of Kirchhotl''s explanation, according to which 

 spots are produced by the absorptive action of clouds 

 floating above the surface of the sun. 



For my own part' I cannot understand how any one who 

 has observed the general phenomena of sun-spots, and 

 especially the peculiarities of appearance they present when 

 near the middle of the disc, can doubt for a moment that 

 they are, or rather indicate, depressions below the surface. 

 There are at least a dozen better reasons for so regarding 

 them in the phenomena of sun-spots, as seen near the 

 middle of the disc, than in the doubtful effects of fore- 

 shortening of spots near the disc's edge. Moreover, the 

 phenomena observed under the former and more favourable 

 conditions would suffice of themselves to render a very 

 flxir account of the peculiarities observed under the latter 



conditions, manifestly unfavourable for deducing exact 

 information. 



I regard the problem to be explained as in reality this : — 

 Since the sun-spots manifestl)' show themselves to be 

 depressions, when observed near the middle of the disc, why 

 do they not always present the appearance due to that form 

 when near the edge? Why in particular should the 

 penumbra on bot/t sides of a spot's shortest diameter when 



Fig. 3.— KincHHoi-F's Explanation or Sun-Spot Changes. 



near the edge be (or seem) often much narrower than we 

 should expect it to be ? 



When we take into account the refractive action of the 

 vapours within the saucer-shaped depres.sion of a .symmetrical 

 spot, we find a sufficient explanation. 



Suppose, for instance, that we have in fig. 4 a rough 

 representation of the section of a spot, when, as yet, the 

 action, whatever it may be (eruptive in my opinion), to 

 which the opening is due, is in its early stage, so that the 

 regularity of the spot is as yet unbroken. The figure shows 

 between spp' s' and a uu' b the streams of ascending glowing 

 vapour, partly condensed into cloud form, which, vertical 

 where there is little disturbance, form the so-called rice 

 grains. We see them near the spot-opening thrust side- 

 ways, so as to be packed near their tops into the gi-eat 

 facular masses seen around a spot. The expansive action 

 of the vapour within the spot region, cai'iied as it has 



Fig. i. 



been from regions of gi-eater pressure deep down below, 

 leads to a cooling (relative only, of course), by which the 

 vapours acquue the absorptive power indicated by the 

 darkening of the spot region. We see also why, looking 

 vertically down into the spot, the central part appears the 

 darkest, for though in looking at the nucleus we are 

 looking towards nn', where the vapour is hotter than at 

 uu', and in looking at the umbra we are looking towards 

 uu', where the vapour is hotter than at pp, which marks the 

 outer limits of the penumbra, yet we have a gi-eater range 

 of absorptive vapours in the first case than in the second, 

 and in the second case than in the thiid. Probably, also, 



