THE GREAT EARTHQUAKE OF S. E. JATAiV 



155 



and making iire-preveutioii almost impossible. It is interesting to 

 observe that the geographical distribution of these fire centres is very 

 similar to that of the centres of seismic intensity, the area crowded 

 by fire centres corresponding to the land badly shaken. It is also to 

 be remarked that there were numerous cases of fire due to chemicals ; 

 the authorities estimate them to have constituted at least 44 out of 

 the total number. 



As to the cause of the rapid growth of the fire (PI. XIX); we 

 must attribute it first to an unfavourable condition of the wind. At 

 the time of the quake, the wind in the city was blowiog nearly from 

 the south with a speed not less than 10 metres per second ; in the 

 evening, it changed its direction to the west and then to the north, 

 attaining at midnight an enormous speed greater than 20 metres per 

 second. This was enough to cause the current of the fire to sweep 

 very rapidly from district to district and back again, and to reduce 

 the whole lower part of the city to ashes. The fatal jtrogress of the 

 wind in the city on those days may be well understood, from Table 

 III, which is the result of observations made at the Central Meteoro- 

 logical Observatory located in the burnt area. Conditions were, however, 

 iXuite ditierent in the surrounding suburbs, where the weather was 

 generally calm. 



Table III. 

 Observations of wind at the Central Met. Obs. 



Sept. 1 



Hour 12 l;] 14 15 1(3 17 18 19 20 21 22 23 



JDirectiou SSW S SSW KW SW 8 W W WNVV NNW NN W NVV 



Speed (in/s; 12.3 11.7 11.3 11.0 10.7 13.7 14.5 13.1 18.1 18.5 19.2 21.8 



Sept. 2 



Hour 0123456789 10 11 



Direction NxWVNiSVV N N N N-SW NNW MW N SSW SVV SW 

 Speed (m/sj 17.6 16.9 15.4 13.8 10.5 10.2 7.9 7.9 5.4 4.G 4.4 5.1 



