93 



Concerning this y. r. it ought to be mentioned that its 

 appearance is a singular one, especially when the slit is a 

 little more opened and the ray is clear. It is then alike a 

 stream of light, not calm, which ordinarily is the case with 

 spectral rays. It happened however often that no light could 

 be seen around the mountain tops or ranges, b ut in the 

 spectroscope the y. r. was observed. The most surprising 

 of all these phenomena was when the y. r. was observed 

 in all directions, as well on the sky as on the earth-surface 

 around the place of observation. It was observed for the 

 first time 21 November 1871 at the parsonage of Enare> 

 Occupied to observe feeble flames along the mountain ran- 

 ges, I was much surprised to see the y. r. whereever I di- 

 rected the spectroscope and with a clearness which did not 

 give room for doubt. From the snow on the soil, from the 

 roof of a barn near at hand or in other words from every 

 place where the spectroscope was directed. The ray had 

 also the above mentioned appearance of a stream of light. 



During the years 1882 — 84 it is noticed that this phe- 

 nomenon was observed 16 times: 1882 the u /xi and 17 /xi5 

 1883 the 26 /i, 1 /ii, 24 /n, 25 /n, 29 /m, 30 /m, 4 /iv, 24 /iv, 6 /ix, 15/ ix. 

 20 /ix, 20 /xi, 21 /xi, 22 /xi *). We will describe the phenomenon 

 as it was observed 20 /xi 1883 in the ofternoon. 



Över the whole sky and on the soil the y. r. was 

 seen though no light could be observed in any direction 

 around the place. Directing the slit against a black wooden- 

 wall, the ray was observed at a distance of 6 m, but va- 

 nished when we approached the wall. It must be concluded 

 that the layer of air itself radiated the light and that this 

 layer ought to be 6 m thick in order to give light enough 

 for the spectroscope. (At an other occasion the y. r. was 

 seen very clearly within a meter from a black cloth placed 

 on a stative). As the spectroscope was directed to a white 



*) The number of these observations would certainly have been 

 greater if a spectroscope had always been at hared on the station, but 

 it was used many times at other places. 



