OBSERVATIONS AT THE SUMMIT OF MONT BLANC. 245 



left below us, we should liave a riglit to infer tliat in rising still more 

 and approacliing the atmospheric limits all the oxygen-rays of the 

 si^ectrum wonld disappear, and that, consequently, the sun has nothing 

 to do with the phenomenon, and contains no oxygen in its atmosphere. 



We come now to the observation made on Mont Blanc on Thursday, 

 the 14th, and Friday, the loth, of September, 1S93. 



We have been able to aver that as a matter of fact the number and 

 enfeebling of the oxygen lines of the solar spectrum seemed to cor- 

 respond to an atmospheric thickness of 4,800 meters, precisely the 

 amount which was underfoot, and that consequently the oxygen lines 

 in the solar spectrum are entirely due to the oxygen of our atmosphere. 



To make sure of this is a delicate matter. On the one hand the 

 heavens must be very pure, and on the other hand a large apparatus 

 must be mounted with care and protected from the wind and from stray 

 light. That is to say, the observation can only be properly made in a 

 closed room sufficiently spacious Those are the conditions which the 

 Mont Blanc Observatory ott'ertd. It was inaugurated by tliat study.' 



This absence of oxygen in the atmospheres of the central orb not 

 merely interests us from the point of view of the future of the worlds 



' What makes tlie novelty of the ohservatious of 1893 is that ou the one hand 

 they have been effected upon the summit of Mont Blanc, and especially that the 

 instrument employed was infinitely superior to that of the two previous ascents. 

 The first was a spectroscope by Duboscq, incapable of separating- the B group into 

 distinct lines, while the second was a spectroscope with a Rowland grating (i)re- 

 sented by Professor Rowland), with telescopes of 0.75 meters focal length, showing, 

 it is needless to say, all known details of B. 



A sjiecial importance attaches to this, because there may be found in the constitu- 

 tion of this B group valuable elements for measuring in some sort the effects of the 

 decrement of the action of our atmosphere as we rise in it, and consequently for 

 .judging whether that decrement corresponds to a total extinction at the limit of the 

 atmosphere. In fact we know that the double lines, whose aggregate constitutes the 

 B group, diminish in intensity as their refraugibility diminishes, or, if you choose, 

 ■with their increased wave length. 



We can avail ourselves of this circumstance, if not to measure, at least to esti- 

 mate the diminution of the action of selective absorption of our atmosphere. In 

 fact, we find that the feeblest doublets vanish .successively in the atmosphere — that is 

 to say, according as the action of absorption diminishes. For example, in ordinary 

 circumstances, on the surface of seas or plains, the maps of the B group show, 

 besides what we call the head of B, thirteen or fourteen doublets. Already at 

 Chamonix, at 1,050 meters, the thirteenth doublet is difficult to ruake sure of. At 

 the Grauds Mulets (3,050 meters) it is only from the tenth to the twelfth that sure 

 observations can be made. At the summit of Mont Blanc I could hardly go beyond 

 the eighth. 



But is it not remarkable that, taking the ratio of atmospheric pressures on Mont 

 Blanc and in the plain, or 



0.43 



-:^ = 0..566, 

 0.76 



and multiplying this by tlie number of doublets generally quite visible in the plain, 

 or say thirteen to fourteen, we get 7.4: or about the number {H) seen by me at the 

 summit ? 



