94 REFOKT — 1895. 



(/) The Observations in Tokio. 



Although tlie observations made in Tokio were carried on at two 

 stations, because these stations were only 1,000 feet apart, and because 

 both were on the alluvium which here forms a layer perhaps 100 feet in 

 thickness above the tuff rock, it was anticipated that the recoi'ds would to 

 some extent be similar in character. For this reason they are described 

 together. 



Machine A, which is similar to those used at Kamakura, is installed on 

 a table-like stone column in my house. The column is 4 feet square and 

 rises clear of the floor from a concrete bed. For a few hours in the 

 morning and in the aftei'noon the sun produced a marked tilting as it 

 shone upon the column through a window on the south side ; on closing 

 this window by a shutter on the outside and by a curtain on the 

 inside, this effect disappeared, while the diagram from a self-record- 

 ing thermometer occasionally showed during 24 hours a steady rise or 

 a steady fall of 1° or 2° C. More usually, however, the diagram showed 

 that from 9 or 10 a.m. until 5 or G p.m. there had been a rise of 4° 

 or 5° C, after which the temperature fell until next morning. This is 

 a point to be noted, because it will be shown that the daily tilting and, 

 in a less marked degree, the intensity of a tremor storm have a similar 

 periodicity. 



The water level beneath my house oscillates above and below 3G 

 feet. 



Machines E and F, which are underground, only differ from A in the 

 fact that their booms are brass tubes. Machines A and F are parallel to 

 each other, and point N.W. Machine E is at right angles to A and F. 

 With an increase in the readings of A or F the u\ovement corresponds to 

 a lifting of the ground on the N.E. side of these instruments. An 

 increase in the scale readings of E corresponds to a tilting on the S.E. 

 .side. 



The underground chamber is excavated on a flat piece of ground about 

 20 feet below the site of my house. It is 13 feet deep and 20 feet square. 

 The floor is covered with \ in. of asphalt, which rests on a bed of concrete 

 6 in. thick, which in turn rests on a bed of well-rammed gravel. Tlie 

 walls and ceiling ai'e brick with clay puddle on the outside. Above the 

 chamber a wooden house has been built ; the entrance is by double doors, 

 and it is fairly well ventilated by gratings for the admission of air and a 

 short iron chimney for its exit. The daily fluctuations in temperature 

 in this underground room are practically zero, the diagram from a self- 

 recording thermometer showing a sti'aight line which at present indicates 

 a steady rising of 1° C. per week. The water level in a well about 80 

 yards distant, where I have established a tide gauge, is about 25 feet below 

 the surface. The floor of the chamber is therefore about 1 2 feet above 

 water level, but it must be remembered that this level may rise and fall 

 through 2 or 3 feet. 



{g) Sensitiveness of tlie Instruments. 



From time to time the sensitiveness of the instruments was determined, 

 and if necessary they were readjusted. 



The first column in the accompanying table indicates the number of 

 millimetres through which the end of the boom travelled by a 1° turn of 



