33 8 IN STARRY REALMS. 



brought it back to the East Indies. The dust of Krakatoa 

 had put a girdle round the earth in thirteen days ! The 

 shape of the cloud appears to have been elongated, so that 

 it took two or three days to complete the passage over 

 any stated place. 



When the dust-cloud had regained the Straits of 

 Sunda the great eruption was over, but the winds were 

 still the same as before, and again the comminuted 

 pumice sped on its impetuous career. The density of the 

 cloud had, however, lessened. Doubtless much of the 

 material was subsiding, and the remainder was becoming 

 diffused over a wider area. Accordingly, we find that the 

 track of the stream during this second revolution is some- 

 what wider than it was on the first, though still mainly 

 confined between the tropics. The speed with which the 

 dust revolved was, however, unabated. Continents and 

 oceans were again swept over with a velocity double that 

 of an express train, and again the earth was surrounded 

 within the fortnight. The dust-cloud had now further 

 widened its limits, but was still distinguishable, and with 

 unlessened speed commenced for a third time to encircle 

 the earth. The limits of the stream had spread themselves 

 outside the tropics, though still falling short of Europe. 

 There is no reason to think that there was any decline in 

 the velocity of 76 miles per hour, but the gradual diffu- 

 sion of the dust begins to obliterate the indications by 

 which its movements could be perceived, so that during, 

 and after, the third circuit the phenomena became so dif- 

 fused that while their glory covered the earth, the distinc- 

 tion between the successive returns had vanished. In 

 November the area which contained the Krakatoa dust 



