ON SEISMOLOGICAL INV'ESTIGATION. 199 



If, however, certain of these instruments have been so arranged that their 

 stability is feeble, or, in other words, so that their free period is large as 

 compared with that of others, they can hardly be expected, even when 

 placed side by side, to yield similar seismograms. A small horizontal 

 pendulum with a period of, say, 50 seconds and a lai'ge multiplication may 

 be continuously in movement over a considei-able period of time. This 

 being the case, it may often happen that the exact commencement of an 

 earthquake may not be determinable. In the Strassburg records, for 

 example, we find commencements of movement given so many minutes in 

 advance of other stations in Europe that for the present, at least, we are 

 inclined to accept the conclusions to which they lead with some reserve 

 (see Earthquake No. 83). 



In Italy there is a great variety of instruments which, for the most 

 part, record with ink upon the surface of paper, or by means of indices 

 writing on smoked paper. 



The ordinary pendulums vary in length from a few metres up to 

 25 metres in length. In Catania, for example, there is a pendulum 

 25 metres in length, carrying a bob of 300 kgs., and with writing indices 

 multiplying its movements 12'5 times. It appears that these exceedingly 

 long pendulums are sometimes affected by the action of the wind upon 

 the building in which they are suspended. When this occurs it becomes 

 difficult to determine with exactness the time at which an earthquake has 

 its commencement. 



The horizontal pendulums are also characterised by their great size. 

 The horizontal booms of such instruments at Ptocca di Papa, which carry 

 25 kgs. near their outer end, are 2-7 metres in length, the tie running to 

 a point 5"25 metres above the foot of each boom. They write with ink on 

 a band of paper moving at a rate of GO cm., or 2 feet, per hour. The open 

 diagrams obtained from both types of instrument are excellent (see 

 p. 207). Unfortunately, the enormous dimensions of these instruments 

 preclude any extensive adoption by private observers. When these 

 dimensions are reduced, as, for example, with the ordinary pendulums, 

 the smaller of these, not having suflicieut multiplication or inertia to over- 

 come the frictional resistance of writing indices, fail in a gi'eater or lesser 

 degree to record the small preliminary tremors, with the result that the 

 time at which an earthquake commences is apparently retarded. 



It is probably sometimes this which explains the great difference in 

 the recorded times at which earthquakes originating at great distances 

 have announced themselves at different recording stations in Italy and 

 Europe. 



For description of instruments in Italy and at Strassburg see 

 pp. 258-272. 



Since writing the Report for 1897 I have obtained a list of records 

 from Japan and the catalogue issued from time to time by Professor 

 Pietro Tacchini in the ' BoUettino della Societa Sismologica Italiana.' Mate- 

 rials extracted from these sources enable me to throw further light upon 

 records published in 1897. 



No. 1, June 15, 1896. (B.A. Keport, 1897.) 



This is the disastrous shock the sea waves accompanying which occa- 

 sioned the loss of nearly 30,000 lives on the N.E. coast of Japan, a 

 description of which will be found in the Report for 1897. 



