318 Intelligence and Miscellaneous Articles. 



lend me his help ; but the tube of 10 metres which we put up for 

 this purpose was not long enough to exhibit the phenomena suf- 

 ficiently. 



I was ultimately able to realize more favourable conditions. A 

 friend of mine, M. Goschler, director of studies at the central School 

 of Architecture, introduced me to the director of the Paris Gas Com- 

 pany, and to M. Arson, chief engineer. These gentlemen placed 

 at my disposal, with a kindness for which I am greatly obliged, the 

 ample resources of this vast establishment. 



An iron tube 37 metres in length was mounted ; it was placed in 

 a wooden box of the same length, filled with well-dried wooden sha- 

 vings, an arrangement which avoids all appreciable loss of heat. The 

 vapour was furnished by a moveable steam-engine of six-horse power, 

 and the light by a lamp of six jets placed in the direction of the axis 

 of the tube. This light, which, as we know, gives a continuous 

 spectrum, enables us to perceive the production of the feeblest dark 

 bands. 



These experiments are now being continued, and I wish simply 

 to communicate to the Academy the first results. These results 

 confirm already, in the completest manner, what the study of the 

 solar spectrum had already indicated. 



In one experiment (August 3, 1866), in which the tube, well freed 

 from air, was full of vapour at the pressure of seven atmospheres, 

 the spectrum was observed with five dark bands, two of which, well 

 marked, spread from D to A (Fraunhofer), and recalled the solar 

 spectrum seen in the same instrument towards sunset. 



From the first comparisons made between the spectrum of aqueous 

 vapour and that of solar light, the group A of Fraunhofer, B (in great 

 part at least), the group C, and two groups between C and D are 

 due to the action of the aqueous vapour of the atmosphere. 



This experiment has further given an interesting result. The 

 spectrum of the transmitted light was seen to be very dark in the 

 most refrangible part, while it was brilliant in the regions of red and 

 of yellow. Thus, although aqueous vapour energetically absorbs 

 certain red and yellow rays, it is very transparent for most of these 

 rays, while it acts in a general manner on the more refrangible rays. 

 It follows from this that aqueous vapour would be of an orange-red 

 colour by transmission, and the redder the greater the thickness 

 through which it acts. 



This result will have to be verified and established with the great- 

 est care, and I only offer it with reserve. If it be finally demonstrated, 

 we shall find in it an explanation of the red colour, so variable in its 

 tints, but always observed at sunset as well as at sunrise. 



The consequences of this discovery of the spectrum of aqueous 

 vapour will be overlooked by no one. We are finally agreed as to 

 the origin of a considerable portion of the rays of the solar spectrum ; 

 and the knowledge of these rays will allow us to investigate the 

 higher regions of our atmosphere as regards their humidity, regions 

 which at present are inaccessible to our means of investigation. But 

 it is more especially in astronomy that the results will be interesting 



