LUMINOUS FLOWERS. 125 



of 1857, 1 repeated the experiment once more. Fancying that 

 the warm weather might possibly have exercised a more than 

 ordinary effect upon the plant, I held a lighted match close 

 to "an open flower, but again without result; in bringing, 

 however, the match close to some other blossoms, it ap- 

 proached a nearly faded one, and suddenly was seen a reddish, 

 crackling, strongly shooting flame, which left a powerful 

 aromatic smell, and did not injure the peduncle. Since then 

 I have repeated the experiment during several seasons ; and 

 even during wet, cold summers it has always succeeded, thus 

 clearly proving that it is not influenced by the state of the 

 weather. In doing so, I observed the following results, 

 which fully explain the phenomenon. On the pedicels and 

 peduncles are a number of minute reddish-brown glands, 

 secreting etheric oil. These glands are but little developed 

 when the flowers begin to open, and they are fully grown 

 shortly after the blossoms begin to fade, shrivelling up when 

 the fruit begins to form. For this reason the experiment 

 can succeed only at that limited period when the flowers are 

 fading. The radius is uninjured, being too green to take 

 fire, and because the flame runs along almost as quick as 

 lightning, becoming extinguished at the top, and diffusing 

 a powerful incense-like smell." 



As to the actual cause of these exhibitions of light, little is 

 known. In the case of M. Fries, the luminosity was always 

 observed between quarter past ten and quarter past eleven 

 in the evening, and especially when the weather was sultry, 

 and was seen to best advantage when the observers were not 

 looking at the flowers directly. The light appeared in fitful 

 flashes, similar to that seen about the other flowers men- 

 tioned, and was supposed to be electric. Mr. Haggeru was 



