PROGRESS OF MKTEOROLOGY IN 1880. 265 



in Jauuary, 1881, the lower limits being practically iiuleteriniiiate. Also, 

 since from Dr. Kiggenbacli's lud Mr. Clark's observations, the glows 

 continued less brilliantly and less prolonged after the first few months 

 right up to the end of 1885, while a decided minimum in the duration, 

 and, therefore, presumably the height, of the reflecting layer, was reached 

 in April, 1881, the important conclusion is arrived at that by that date 

 the larger and more effectively reflecting particles had descended to a 

 lower level, leaving the finest particles suspended at nearly the same 

 elevation as at first. This is further corroborated by the remarkable 

 fact that the large corona reached its maximum intensity during the 

 same month. 



Finally, in section vii, Mr. Archibald gives a general analysis of the 

 connection between all the optical phenomena and the eruptions of Kra- 

 katoa, both in May and August, in which the various objections on the 

 ground of the initially rapid transmission of the' appearances, insuffi- 

 ciency of fine, solid ejecta, length of time of its suspension, and the 

 occurrence of apparently similar phenomena on dates previous to the 

 great August eruption are discussed in turn. The time of suspension 

 of the finest dust in particular is shown — by an application of Profes- 



sor Stokes's formula, V=^'l( 1 Vj^, for the velocity of a small par- 

 ticle descending in air, and in which viscosity is properly considered — 

 to be over two years between 50,000 and 100,000 feet, even assuming 

 the particles to be spherical, which is the most unfav^orable supposition. 

 If, as is most probable, they wee thin plates, the time would be much 

 longer. A final summary is then given of the direct and local connec- 

 tion between the optical i)henomena and the eruptions, both of May 

 and August, which the subsequent discovery of the relative though 

 minor importance of the May eruption rendered necessary. 



Sunset glows. — Prof. Cleveland Abbe has published a paper on the 

 sunset glows of 1884-'85 (written in November, 1885), in which he 

 shows that the phenomena can not be produced by refraction and con- 

 sequent dispersion through small drops, but are explicable only as 

 diffraction effects in which the nature of the substance, whether minute 

 drops of water or non- transparent particles of dust, is immaterial. 



The Bishop's Ring is attributed to particles so far removed from the 

 earth's surface as to remain sensibly permanent through many seasons, 

 while the red twilights are diffraction rings due to similar and slightly 

 larger particles in the lower atmosphere. In a prefatory note written 

 February, 1889, Prufessor Abbe gives the following as his present con- 

 clusions : 



(1) Vapor haze is more important than dust haze. 



(2) A shallow layer, sparsely filled with such minute particles of 

 vapor haze generally accompanies every area of high pressure and clear 

 air, and appears to produce the diffraction necessary for the phenomena 

 that are still observable. 



