ON SEISMOLOGICAL INVESTIGATION. 229 
Casamicciola revealed oscillations of 10”, due probably to the earthquake 
which caused the great sea-waves in Japan ; but, at the same time, the 
levels were not affected, though the corresponding traces on their records 
would have been 2:5 mm. in length. 
The most marked diagram so far obtained is that which was due to the 
severe earthquake in Carniola on April 14, 1895 (see fig. 18, in which the 
scale is half that of the original diagram). The curve in this case was, 
however, obtained before the employment of smoked paper, so that it is 
not so clear as that of a similar earthquake at the present time 
would be. 
APPENDIX. 
Notes on SpeciaL Eartuquakes. Ly Professor J. MILnNe. 
About 11.30 p.m. on August 26, 1896, a diagram was obtained which 
may represent an earthquake that occurred in Iceland on that date. It 
shows three maxima. A much more remarkable record, however, is one 
commencing as a series of minute tremors at about 8.23 a.m. on August 31. 
Fie. 19.—Japan Earthquake ; Carisbrooke Castle Record. 
It is shown on the photogram from Shide, and also from that at Caris- 
brooke Castle (fig. 19), and the times of marked phases of motion in G.M.T. 
are as follows :— 
| Carisbrooke 5 
ate Castle Shide 
a STS 
eG Se |f HM, S. 
1. Exceedingly minute tremors, August 30 . 20 23 6 | Too faint to be visible. 
2. First decided tremors : : 4 . 20 31 46 | 20 31 42 
83. Heavy motion commences. : : . | 20 57 6 | 20 56 49 
4. First maximum (about) . . : e021 4526 | 2 od> Oo 
beeliemaximom jf) 5... . + | 21 14 26 | Not calculated. 
6. Heavy motion . : , , : Ses ON AG oe * 
ls s 43 - 5 ; ; F . | 21 23 6 | 21 24 43 
Be. 9 ” : 3 : : : . | 21 27 46 | Not calculated. 
9. End of tremors . : , : 3 . | 23 16 20 | 22 59 36 
Duration of disturbance . : ; . | 2 63 20 — 
Duration of preliminary tremors . .| 0 34 0 — 
