224 REPORT — 1887. 



drawn as curves, the ordinates of which represented the amplittide of 

 tremor motion. These carves were compared with curves which repre- 

 sented the force of the wind and the height of the barometer in Tokio. 

 After comparing these curves with each other it appeared that the micro- 

 seismical distm-bances showed the most varying relationship with the 

 strength of the wind and the movements of the barometer. 



(h) Earth-tremors and earthquakes . — Professor M. S. de Rosse has 

 pointed out some remarkable instances when earth-tremors have been the 

 precursors of earthquakes. From my records it appears that earthquakes 

 have happened fifty-three times when there were no tremors, and thirty- 

 three times when there were tremors. From this I conclude that earth- 

 quakes are just as likely to occur without tremors as with them. 



(i) Earth-tremors and the state of the ivind in Central Japan in 1885. — 

 Central Japan is here meant to include all places within about 200 miles 

 of Tokio. In this area there are eleven meteorological stations. If the 

 wind has had a force of three or upwards at more than one of these 

 stations it has been considered windy. When the wind has not exceeded 

 two or one, even if wind of that intensity was blowing at all the eleven 

 stations, it has been considered calm. An arbitrary division of days or 

 periods into windy and calm, such as has here been followed, must neces- 

 sarily result in absolutely separating the days which were truly windy 

 from those which were truly calm. There are, however, a number of 

 cases which might equally well be placed in either group. 



In 1885 there were 945 weather maps which could be compared with 

 the records of the tremor instrument. The results of the comparison were 

 as follows : — 



1. With no wind and no tremors there were . . .651 cases. 



2. With no wind and tremors „ ,, . . .51 



3. With wind and no tremors ,, >, • • .60 



4. With wind and tremors „ ,, . . .65 



5. With a local wind in Tokio and no tremors there were 101 



6. With a local wind and tremors there were . . .17 



On the assumption that tremors are due to the wind, then the second 

 and third results are difficult to understand. These have therefore been 

 carefully re-examined, with results as follows: In 17 out of the 51 cases 

 of tremors occurring when there was no wind it is found that at these 

 times it was moderately windy, and it is therefore possible that the tremors 

 which were observed might have been due to wind. In 8 instances the 

 tremors were accompanied by marked barometrical depressions, while in 

 the 26 remaining cases the tremors were slight and of short duration. 



In 51 cases out of the 60 cases when there was wind and no tremors 

 it is seen that the wind was only moderate and of short duration. Most 

 of these winds were afternoon sea-breezes, which possibly do not continue 

 sufficiently long to produce tremors. In 9 instances of tremors they are 

 the result of wind. These tremors ought to have been observed. The 945 

 comparisons may therefore be arranged as follows : — 



1. With no wind and no tremors ..... 651 cases. 



2. With no wind and tremors 51 cases : — 



(a) Tremors possibly due to preceding wind . . 17 „ 



(6) Tremors accompanying barometric depressions . 8 „ 

 (c) Tremors possibly of subterranean origin . . . 26 „ 



