I 



M. Arago on the Physical Constitution of the Sun. 203 



of the results. But astronomers exercise no influence over 

 the phenomena which they study ; they are obliged to wait 

 sometimes for centuries until the celestial bodies present 

 themselves in a favourable position for the resolution of a 

 difficulty. -ujuM'i" 



On this occasion, however, the doubts raised by the obser- 

 vations of 1842 have already, in the course of last year, been 

 subjected to a new experimental examination. An eclipse 

 of the sun was announced to occur on the 8th of August 

 1850, which was to be total in the Sandwich Islands. 



The captain of the ship Bonnard, commanding our station 

 of Otaheite, entertained the happy thought of sending M. 

 Kutscki, superintendent of roads and bridges, to Honolulu, 

 in the island of Taheite, the capital of the Sandwich Archi- 

 pelago. 



The account which we have received from this skilful ob- 

 server, contains the following sentence : — " The slender and 

 reddish striated appearance which was found near the north- 

 ern prominence seemed to be completely detached from the 

 margin of the moon." 



Still later, in the eclipse of 28th July 1851, Messrs Mau- 

 vais and Goujon of Dantzic, and the celebrated foreign as- 

 tronomers who had repaired to different parts of Norway, 

 Sweden, and the north of Germany, saw in all the selected 

 stations without exception, a spot uniformly red and sepa- 

 rated from the limb of the moon. 



The observation of M. Kutscki, and the corresponding ob- 

 servations of 1851, put a definite termination to the expla- 

 nations of the protuberances, founded on the supposition 

 that there existed in the sun mountains whose summits 

 would reach considerably above the photosphere. 



When it shall have been clearly demonstrated that these 

 luminous phenomena cannot be the effect of the inflexions 

 which the solar rays might experience in passing near the 

 rough parts which fringe the circumference of the moon ; 

 when it shall have been demonstrated that these rosy tints 

 cannot be assimilated to simple optical appearances, and 

 have, in truth, a real existence, that they are not real solar 

 clouds, it will then be necessary to add a new atmosphere to 



