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



speaking. Explanations giving neither an exact account of 

 the height, the form, the colour, nor the fixity of a phenome- 

 non, ought to have no place in science. 



Let us come to the idea, much extolled for a short time, 

 that the protuberances of 1842 were solar mountains, whose 

 summits extended beyond the photosphere covered by the 

 moon at the moment of observation. 



Following the most moderate computations, the elevation 

 above the solar disc of one of these summits, would have 

 been 19,000 leagues. I am well aware that no argument, 

 because based on the vastness of this height, should lead to 

 the rejection of the hypothesis. But it may be much shaken 

 by remarking, that these pretended mountains exhibit consi- 

 derable portions beyond the perpendicular, which, conse- 

 quently, in virtue of the solar attraction, must have fallen 

 down. 



Let us take a rapid glance of a fourth hypothesis, accord- 

 ing to which the protuberances would be assimilated to solar 

 clouds floating in a gaseous atmosphere. 



Here we find no principle of natural philosophy to prevent 

 our admitting the existence of cloudy masses from 70,000 

 to 90,000 miles in length, with their outline serrated, and 

 assuming the most distorted forms. Only in further pur- 

 suing this hypothesis, one could not fail to be astonished that 

 no solar cloud had ever been seen entirely separated from 

 the circumference of the moon. 



It is towards this determination, the subject otherwise 

 eluding us, that the researches of astronomers should be di- 

 rected. 



A mountain being incapable of sustaining itself without a 

 base, the fortuitous observation of a prominence, separated 

 in appearance from the margin of the moon, and, conse- 

 quently, from the real margin of the solar photosphere, should 

 be sufficient utterly to overthrow the hypothesis of solar 

 mountains. 



But let it be distinctly remarked, that it is not with astro- 

 nomical researches as with those of chemists and mechanical 

 philosophers ; these latter can at will vary the conditions 

 under which they operate, and which may change the nature 



