82 Barus — Geometric Sequences of the Coronas of Cloudy 



too fleeting, and if obtained artificially with nuclei, the colors 

 are apt to be too dull for good discrimination. The charge of 

 nuclei is removed too rapidly by the condensation ; and if the 

 tube is horizontal, two longitudinal vortices are seen immedi- 

 ately after exhaustion, the air rising on the outside and descend- 

 ing into the center. Finally there is no guaranty that the 

 charge of nuclei added is uniformly distributed. Tubes have 

 one advantage, however, of affording a demonstration of the 

 occurrence of rigorously axial colors. 



For studying coronal effects spherical receivers, 30 cm. or 

 more in diameter, as used by Coulier and Kiessling, are prefer- 

 able. With divergent sunlight the display is gorgeous, the 

 colors glowing metallically. The mantel of a Welsbach burner 

 seen through a small hole in a screen is better for experimental 

 purposes. The axial colors are not very vivid from the small 

 thickness of medium traversed ; but the diffraction colors are 

 bright, particularly on blotting out the central beam with a 

 small black circular screen. Vortices are no longer violent. 

 They serve a useful purpose in keeping the contents of the 

 receiver homogeneous. 



As a whole, the defraction pattern is a contraction inward of 

 dark rings as the particles, during the course of exhaustion of 

 the nucleated moist air, grow larger. With white light these 

 occurrences are, however, by no means a succession of ordinary 

 coronas. The initial and particularly brilliant coronas corre- 

 sponding to finer particles have colored* central fields, and it 

 is only after many successive exhaustions that the normal, truly 

 white centered corona is reached. 



The information first to be sought is some classification of 

 the sequence of coronal colors and of the axial colors seen in 

 these experiments. These contrasting phenomena are radically 

 different. If the former are in large measure diffractions, the 

 latter are absorptions. The axial color is always nearly compli- 

 mentary to the central field of the corona. As the particles 

 grow in size, both central color phenomena pass through New- 

 ton's series at a definite phase difference apart, but the axial 

 colors grow faint and vanish long before the diffraction colors. 



Coronas. 



3. Loss of nuclei by exhaustion. — Omitting preliminary 

 results, it will now be necessary to investigate data of a quanti- 

 tative character, serving to distribute the coronas in a scale of 

 decreasing numbers of nuclei. It will not be feasible to arrive 

 at the latter datum at once ; for the number of nuclei must be 



* The narrow white sheen around the axis mentioned in the tables below is 

 probably irregularly diffused light. 



