122 REPORT— 1895. 



During the day, or from 7 A.M. until about 9 p.m., tremors are absent, but 

 they occur in a very marked manner between 9 p.i\r. and 7 a.m., when they 

 suddenly cease (tig. 9). 



The central part of fig. 3 shows a white band the width of a small plate 

 of blackened mica at the end of a reed boom. In the mica there is a 

 broad and a narrow slit, which correspond to the broad and thin lines in 

 the diagram. The object of the broad line is to obtain a photogram, 

 when the boom and its plate of mica are displaced rapidly, at which times 

 sufficient light may not pass through the narrow slit to alFect the bromide 

 paper. When the motion is slow the tine slit gives the best definition. The 

 vertical white marks at distances of 41 mm. apai't are made by the pro- 

 longation of the minute hand of a watch crossing one end of the fixed 

 slit every hour. If the instrument be disturbed at knoton times, and the 

 times at which these disturbances took place be determined by the 

 irregularities produced on the broad and thin bands, the errors in these 

 readings vary between three and twenty seconds. Should the paper 

 move irregularly, this is shown by ditferences in the length of the spaces 

 representing hours, while the times at which retardation or acceleration 

 took place are shown by vertical striations in the broad black bands. 

 This is the form of recording surface which I am using to obtain photo- 

 grams of movements due to earthquakes. It is not suitable for tremors 

 or daily waves. 



Pendulum at S. — This pendulum, originally at N, commenced its 

 records on April 24, 1895, at the Agricultural School at Komaba, which 

 is five and a half miles distant in a S.W. direction from the University. 

 Komaba is .situated on a flat plateau, and the nearest irregularity in the 

 contour of the ground from the instrument which stands in the middle of 

 a field partly covered with corn is at a distance of about half a mile, where 

 there is a small valley. The soil is so light and dusty that a walking- 

 stick may be pushed into it for a depth of two or three feet. Beneath 

 this comes a red earth. The records show tremors, but they are small. 



A westerly motion of about 5 mm. is completed at about eighteen 

 hours, or 6 A.M., and an easterly movement of equal amount at about three 

 hours, or 3 p.m. (fig. 11, p. 137). 



On the S.E. side of the instrument the ground is bare, and during the 

 day the pendulum moves to this side. 



(6) Daily Wave Records. 



In the following tables the localities or instruments are indicated by 

 capital letters. The times at which an instrument completed its N.E., 

 S.W., E., or W. excursion are indicated by hours, or 24, corresponding 

 to mid-day, and 12 to midnight. The figui'es in brackets give in milli- 

 metres the amplitude of the displacements. When a dash takes the place 

 of hours it means that the displacements were too small to be measured, 

 or that the diagram was a straight line. When the space for hours is 

 left blank, it signifies that for some reason or other, either that no diagram 

 was taken or that it has been lost. A record like 9-15 on October 15 

 means that the S.W. motion of A was completed between nine and fifteen 

 hours. 



