DECOMPOSITION BY LIGHT 115 



right and left, issued an axis uniting the globe with two 

 adjacent cylinders. Both globe and cylinders were ani- 

 mated by a common motion of rotation. As the action 

 continued, paroxysms of motion were manifested; the vari- 

 ous parts of the cloud would rush through each other with 

 sudden violence. During these motions beautiful and gro 

 tesque cloud-forms were developed. At some places the 

 nebulous mass would become ribbed so as to resemble the 

 graining of wood; a longitudinal motion would at times 

 generate in it a series of curved transverse bands, the 

 retarding influence of the sides of the tube causing an ap- 

 pearance resembling, on a small scale, the dirt-bands of 

 the Mer de Glace. In the anterior portion of the tube 

 those sudden commotions were most intense; here buds 

 of cloud would sprout forth, and grow in a few seconds 

 into perfect flower-like forms. The cloud of iodide of 

 isopropyl had a character of its own, and differed materi- 

 ally from all others that I had seen. A gorgeous mauve 

 color was observed in the last twelve inches of the tube; 

 the vapor of iodine was present, and it may have been 

 the sky-blue scattered by the precipitated particles which, 

 mingling with the purple of the iodine, produced the 

 mauve. As in all other cases here adduced, the effects 

 were proved to be due to the light; they never occurred 

 in darkness. 



The forms assumed by some of those actinic clouds, as 

 I propose to call them, in consequence of rotations and 

 other motions, due to differences of temperature, are per- 

 fectly astounding. I content myself here with a meagre 

 description of one more of them. 



The tube being filled with the sensitive mixture, the 

 beam was sent through it, the lens at the same time being 



