400 PROFESSOR C. MICHIE SMITH ON OBSERVATIONS ON 



Hicks Pasha, dated Duem, September 24th — " To-day, when it [the sun] rose 

 it was of a pale green colour." There is no proof, of course, that this is the 

 date of its first appearance in the Soudan, but if the green was nearly as 

 brilliant there as it was here, it could not have escaped observation. 



Before discussing the causes of the phenomena that have been described, 

 it will be well to point out that we must distinctly separate the green sun and 

 the sunsets that appeared along with it from the remarkable sunsets which 

 occurred nearly all over the world some time later, and which were visible 

 here till at least the end of April. My reasons for making this separation are 

 — 1. The general appearance was quite different in the two cases. The sunsets 

 accompanying the green sun were peculiarly lurid, and were, so far as I could 

 judge, simply an exaggerated form of what we usually have some time before 

 the bursting of the monsoon, and round the horizon we had then a most 

 decided fog, in which stars were lost some time before setting. The main 

 features of the subsequent sunsets, on the other hand, were the delicacy of the 

 colours, and the beautiful rosy afterglow reflected apparently from very light 

 cirrus clouds high up. The horizon at this time was remarkably clear, as is 

 shown by the circumstance that Mr Pogson was able to make accurate measure- 

 ments of the faint comet " Ross " within less than 5° of the horizon. 2. The 

 spectrum was totally different in the two cases, for in the latter case the red 

 end was very free from general absorption A, a, and B standing out clear and 

 sharp, while the rain-band was slight, or altogether wanting, and the low-sun 

 bands were strong — a complete contrast to what I have described as being 

 visible in the former case. I may add, that any increase in the strength of the 

 rain-band has been accompanied by a decrease in the brilliance of the after- 

 glow. 



To account for the green sun, three hypotheses have been put for- 

 ward — 



1. That it was due to vapours or dust from the volcanic eruption at 

 Krakatao. First suggested by Mr Pogson. 



2. That the cause was the presence of an abnormal amount of aqueous 

 vapour. An explanation which I offered at the time of the occurrence. 



3. That it was caused by a cloud of meteoric dust. 



Mr Lockyer has taken up the first of these theories, and starting with the 

 assumption that the sunset effects were due to the same cause as the green 

 sun, has attempted to trace the general propagation of the dust round the 

 earth. His first line passes through Mauritius, the Sechelles, Cape Coast 

 Castle, Brazil, Trinidad, Panama, and Honolulu. His second is a line passing 

 from south to north through India. Taking the dates of the appearance of the 

 phenomena which can be relied on, we find that they would require approxi- 

 mately the following velocities of propagation — 



