310 



REPORT — 1880. 



of the atmosphere, the nature of the collodion, and the duration of the 

 exposure ; but, if the finest days be chosen, and a collodion of constant 

 composition be used, the length of the exposure being always the same, 

 then comparable series are obtained. For example — 



Observations made at Courtenay (Loiret), 11th Sept. 1878. Lat. 48° 

 2' 2" ; wet plates. 



The length of the spectrum is expressed in wave-length by compari- 

 son with the diagram constructed by M. Cornu from observations made 

 in 1877.' The table given above indicates that the length of the spectrum 

 diminishes with the height of the sun ; which tends to prove, M. Cornu 

 thinks, that atmospheric absorption is the cause of the limitation. 



Therefore, on diminishing the depth of the atmosphere, by making the 

 observation at a greater elevation, the length of the photographic spectrum 

 ought to be increased. From calculations based on experiments made at 

 a small elevation M. Cornu concludes that the limit of the spectrum 

 would be altered by an amount equivalent to one-millionth of a milli- 

 metre for a rise of about 663 metres. He deduces from his experiments 

 a formula which expresses the law governing the increase in length of 

 the specti'um at different heights. This formula shows, says M. Cornu, 

 that if the principles which have served to establish it are exact, we are 

 condemned never to know a considerable portion of the spectrum — per- 

 haps the most interesting. 



The spectrum of iron in the voltaic arc extends at least to radiations 

 of which the wave-length X=:200, whereas the most favourable obser- 

 vations of the solar spectrum have only reached X = 293. 



The consideration of the temperature of these two sources leads one to 

 think that the solar spectrum ought to extend beyond the limit of the 

 spectrum of the voltaic are. 



From theoretical considerations, based on these observations and the 

 formuliB deduced from them, M. Comu concludes that if this absorption 

 really takes place it ought to be sensible for small thicknesses of atmo- 

 sphere ; in other words, there ought to exist rays of so short wave-length 

 as to be absorbable by a small thickness of air.^ He finds that with 

 the solar rays approximately the following results should be obtained : — 



lO'OOra. of air at 760mm. would extinguish rays the wave- "1 qii .04 



length of which is ,..../ 



1-OOm. „ „ „ „ „ „ . 184-21 



O-lOm. „ „ „ „ „ „ . 156-58 



The sparks produced by a powerful induction coil approximately fulfil 

 the double function of furnishing very intense and very refrangible 

 radiations. This is particularly the case with the extreme rays of alumi- 

 nium, designated 30, 31, and 32 by M. Soret, and the wave-lengths of 

 which were found by M. Cornu to be : — 



Ray 30 . . . 

 „ 31 double 



198-81 

 / 193-35 strong. 

 1^192-87 weak. 



r 186-02 strong. 

 32 triple < „ very weak. 



I 185-22 not so strong. 



• Comjates Rendus, t. Ixxxvi. 



2 Ibid. t. Ixxxviii. 



