80 REPORTS ON THE STATE OF SCIENCE. 



after the discovery of the two Americas. To this we may add that it has 

 only been of late years that the records of certain countries with an ancient 

 civilisation have become accessible to us. Up to the year 18.50 Europe 

 imagined that in Japan there were from one to six earthquakes per year. 

 Now we know there are at least 1,000. We also know that the Japanese 

 have for a long time past published ' Jishin-Nendaiki,' or earthquake 

 calendars, which give records of very destructive earthquakes since 

 extremely early times. 



As an illustration of the difference between the new catalogue and 

 the one published by Mallet, who endeavoured to exclude from his 

 registers very small disturbances — such as ' after-shocks ' — I find that 

 between ISOO and 1808, which are yeai-s taken at random. Mallet 

 gives 407 entries, but of these only thirty-seven are stated to have pro- 

 duced structural damage. All other large catalogues appear to be of the 

 same nature. Large earthquakes which have announced changes of geo- 

 logical importance in the earth's crust stand side by side with great 

 numbers of seismic trivialities, many of which may not even have rattled 

 a window. To give each unit in such collections an equal value, and this 

 has frequently been the case, would not for many analyses lead to satis- 

 factoi'y results. For example, the old catalogues when taken en bloc, 

 could hardly be expected to give us accurate information about the distri- 

 bution of seismic energy in time. 



The sources from which material has been obtained to form the present 

 catalogue are various. From a.d. lb to 1842 I am indebted to Mallet's 

 catalogues, from that date to 1864 I have in great measure relied upon 

 Alex. Perrey. Some of his writings, however, appear to be only in MS., 

 and therefore have not been accessible. To obtain records of the last 

 forty-three years, however, has been a matter of considerable difficulty. 

 Appeals to daily journals like the ' Times ' have proved to be very disap- 

 pointing. For example, the records of large earthquakes in the 'Times ' 

 fur 1855 and 1864 were respectively five and one, while references to 

 foreign journals and the Transactions of learned societies show that verv 

 destructive earthquakes occurred in each of those years from twenty- seven 

 to thirty times. 



If this catalogue proves to have more homogeneity in its character 

 thaa its predecessors, it is to be hoped that analyses of the same will 

 lead to more definite results than have hitherto been obtained from 

 earthquake statistics. 



An addition to this work is a Catalogue of Chinese Earthquakes. It 

 has been compiled by my assistant, Mr. Shinobu Hirota, and, so far as I 

 am aware, it is the most important collection of records relating to 

 earthquakes in China which have ajDpeared in a European language. 



As an illustration, and al.so as an introduction to analyses which 

 may be made from a catalogue of destructive earthquakes, I give the 

 following : — 



Oyi the Seismicity in Europe and adjacent Countries between the Years 



A.D. 1000 and 1850. 



The subjoined table shows the number of very destructive earth- 

 quakes which have been recorded between a.d. 1000 and 1850. The 

 grouping is in periods of fifty years. Under the heading marked III. 

 we have the number of exceedingly destructive earthquakes, the records 



