330 IN STARRY REALMS. 



Greenwich mean time. This is probably correct within 

 two or three minutes. Diverging from its source this 

 wave reached Greenwich after an interval of a little more 

 than ten hours. The interesting point is, however, the 

 determination of the period of a complete oscillation, that 

 is to say, the interval between the passage of the wave 

 over Greenwich and the next passage of the wave in the 

 same direction also over Greenwich. It has been found 

 convenient to designate the successive waves as i., ii., iii., 

 iv., &c., the odd numbers being those from Krakatoa to its 

 antipodes, and the even numbers being the return waves 

 from the antipodes to Krakatoa. At Greenwich, for 

 example, we find the interval between i. and iii. to have 

 been 36'47 hours, between iii. and v. 36'82 hours, and 

 between v. and vii. 37*05 hours. For the return waves 

 the intervals between ii. and iv. was 34'78 hours, and 

 between iv. and vi. 35 '25 hours. The similar values vary 

 slightly when obtained at the several stations, but the 

 average results indicate that for its first circuit of the 

 earth the wave required 36 hrs. 24 mins., for the second 

 36 hrs. 30 mins., and for the third 36 hrs. 50 mins. The 

 similar periods for the waves travelling in the reverse way 

 were 34 hrs. 46 mins., and 35 hrs. 4 mins. respectively. 

 The average of all is very nearly a day and a half. 



Before leaving this part of the subject, I must refer to 

 the approximate identity between the velocity of this 

 aerial disturbance and the velocity of ordinary sound. 

 This is well brought out by General Strachey. The speed of 

 the wave varied from 674 to 726 miles per hour. The 

 speed of sound propagation is 723 miles at zero Fahren- 

 heit, and is 781 miles at 80 Fahrenheit. Considering 



