448 MR WILLIAM SWAN ON THE 
ing motion. Its base was from first to last sharply bounded by the edge of the 
moon.” ‘To my great astonishment,” he adds, “this marvellous object con- 
tinued visible for about five seconds, as nearly as I could judge, after the sun began 
to reappear, which took place many degrees to the south of the situation it oc- 
cupied on the moon’s circumference. It then rapidly faded away, but it did not 
vanish instantaneously.’”* 
These observations seem quite inexplicable, on the hypothesis that the pro- 
minences result from mirage occasioned by the unequal heating of the air. For 
not only did they preserve their forms unchanged during a period at which little 
or no unequal heating of the air could have taken place; but according to the 
very important observations of Mr Dawes and Mr Htnp, they continued visible, 
apparently without change of form, even after the reappearance of the sun. Now, 
at the reappearance of the sun, the air in the path of light would rapidly pass 
through the three states, of being first entirely protected from the sun’s rays, 
then heated on one side at the moment of reappearance, and finally heated on 
both sides.t About that time, then, if phenomena of the nature of mirage ex- 
isted, we might expect the most rapid and conspicuous changes of form to occur ; 
but instead of this being the case, the prominences retained their forms unal- 
tered, until they vanished before the direct light of the sun. On these grounds, 
we must therefore regard the hypothesis which would refer them to the unequal 
heating of the air, as quite untenable. 
* Ast. Soc. Notice, p. 69; or Astronomische Nachrichten, No. 777. 
M. Mayerte at the eclipse of 1842, saw one of the red prominences after the sun had reappeared 
(quelques instantes apres emersion du Soleil.)—Annuaire, for 1846, p. 411; see also p. 421. M. 
Conti saw the prominences for a long time (per lungo tempo), after the reappearance of the sun ; and 
M. Bieta for some seconds, pp. 428,429. The statement of the latter observer is particularly 
explicit. ‘“ Les premiers rayons dv Soleil se montrerent en divers points séparés. Bientét ces points 
se réunirents et formerent une lunule tres-déliée. Quelques secondes aprés la formation de cette lunule, 
les pyramides rougeatres cessérent de se voir.” 
+ May not the unequal heating of the air on the two sides of the path of the solar rays be the 
chief cause of the remarkable fluctuations in the sun’s light, which have been observed at the be- 
ginning and end of the total phase of a solar eclipse? M. Savournry, an observer of the eclipse of 
July 1842, relates, “ On a vu ici des ombres et des taches lumineuses courir les unes aprés les 
autres, comme paraissent le faire les ombres produites par de petits nuages qui passent successivement 
sur le Soleil. Ces taches n’étaient pas de Ja méme couleur; il y en avait de rouges, de jaunes, 
de bleues, de blanches. Les enfants les poursuivaient et essayaient de mettre la main dessus. Ce 
phénoméne extraordinaire fut remarqué quelques instants seulement avant la disparition compléte 
du Soleil.’—Annuaire for 1846, p. 893. The strata of illuminated and dark air at the surface of 
the moon’s shadow, if their temperatures, and consequently their densities differ, cannot fail to 
mingle irregularly, and occasion fluctuating movements in the transmitted rays of light, similar to 
those which cause the dancing motion of objects seen through an ascending current of heated air, 
or through liquids of unequal densities which are in the act of mixing. This may also serve to ex- 
plain the flickering appearance of the prominences noticed by some observers ; which, from the terms 
used in deseribing it, was evidently not a permanent change of outline, but merely a fluctuation 
of their forms about a mean condition. Thus Mr Dawrs and Mr Goon, who saw on the moon’s 
southern limb a long range of low prominences, both describe it as in motion. Mr Dawes, however, 
says, its irregularities appeared permanent, and he aseribes its undulation to our own atmosphere. 

