METEOROLOGY. 381 



certainly never around the world, and to all latitudes from GO north to 

 40 south ; neither does it ever give rise to anything like the brilliant 

 purple twilight. The great cloid of smoke due to the forest fires of 1871 

 (not to the Chicago tire as erroneously assumed) was observed by navi- 

 gators in the mid Atlantic. The haze is in fact composed of minute 

 particles of carbon, each of which, as is well known, has the property of 

 condensing upon itself a small atmosphere of other gases, including 

 aqueous vapor. This qnasi-chemical action seems to hold the vapor in 

 an invisible state at a lower temperature than would have been the case 

 were no particles there; it hinders condensation so that rain and cloud 

 are lessened during the prevalence of this smoke. When however rain 

 does form, it falls with unusual severity. It is, we think, likely that to 

 this property of carbon or to the special radiating and cooling proper- 

 ties of fine solid i)articles that we must attribute the phenomena dis- 

 cussed b}' Aitken, as there can be no doubt that aqueous vapor will con- 

 dense into fog particles and larger drops even in an atmosphere free 

 from fog and solid particles.] 



387. [The remarkable red sunsets and sunrises of October to Decem- 

 ber, 1883, have formed the subject of innumerable contributions both 

 from an observational and theoretical standpoint. Exhaustive investi- 

 gations are promised by committees of societies, especially the Royal So- 

 ciety, which have undertaken to collect all the data relative to these 

 important phenomena. The hypothesis that these were due to the 

 gradual si)read over the northern hemisphere of dust ejected from the 

 tremendous eruption of Krakatoa, started by Bishop, September 5, 1883, 

 and Lockyer, October, 1883, was first modified into the assumption that 

 moisture as well as, or even instead of, dust was more plausible ; the 

 possibility being granted that the vapor exists in the form of minutest 

 spiculai of ice or of minute spheres. Subsequently it became a query 

 whether circumstances favorable to the dissemination of ice spiculie or 

 vapor in the higher regions might not have existed independent of the 

 Krakatoa eruption ; this view was then strengthened by the discovery 

 that similar twilight phenomena had been observed some time before 

 the Krakatoa eruption, while on the other hand it was also shown 

 that several volcanic eruptions in past years had been followed by 

 similar sky colors. The renewal of the phenomena in 1884 made it 

 highly i)robable that we have to seek the cause of these colors in some- 

 thing outside of the earth's atmosphere ; during a whole year the sun 

 liad been surrounded by a large irregular cloud of haze, extending 20 

 to 40 degrees from its center in all directions. This haze was decidedly 

 streaky or sti^ated, and occasionally showed a pinkish tint. At })resent 

 the most plausible hypothesis seems to be that the earth in the autumn 

 of 1883, and again of 1884, passed through a stream of gaseous or va- 

 porous meteoric matter, which temporarily combined with the earth's 

 atmosphere, while the great mass passed on to the sun, and has accu- 

 mulated as a nebulous cloud about that body. The color phenomena, 



