456 Scientific Proceedings, Royal Dublin Society. 



pied themselves with this question ; and De Eossi, in a very remark- 

 able address, delivered before the Greographical Society of Eome, in 

 February, 1882, oi; his then projected seismical and endodynami- 

 cal map of Italy, enters into very interesting considerations on this 

 point, to some of which I shall take occasion to call attention. 



One of the most frequently-cited observations in connexion 

 with earthquakes, at least within the last century or so, is the state 

 of the barometer before and after the shock. It is frequently re- 

 corded that the barometer was low before the earthquake cited, and 

 a sudden fall has come to be construed, at least for certain locali- 

 ties, as an indication of an impending shock. Such an interpreta- 

 tion of the fall necessarily implies a connexion or relation between 

 the atmospheric pressure at the place considered and the state of 

 equilibrium of the tensions existing in the constituent matter of 

 the earth underneath that place. Are there any grounds'for assum- 

 ing the general existence of such a relation ? I believe there are. 

 The atmospheric pressure at any one place is simply the measure of 

 the attraction exercised by gravity on the atmosphere at the place ; 

 that is, by the mass of the earth. Any variation in the distribu- 

 tion or arrangement of this must in some sort alter the position of 

 the centre of gravity of the earth, however minutely, and should 

 have, as consequence, a corresponding alteration in the amount 

 and direction of the attraction exercised on the atmosphere, and 

 therefore in the atmospheric pressure over different parts of the 

 surface of the earth. Now, tensions which exist under a given 

 surface, and which increasing and tending to reach a limit, and 

 therefore of a nature capable of giving rise to an earthquake, must 

 affect the arrangement of the matter of the earth under that sur- 

 face, and pro tanto are capable of affecting the attraction of the 

 earth for the atmosphere at the place considered, and consequently 

 the pressures. 



But, even admitting this relation, the difficulty lies in the 

 measurement of the amount of pressure attributable solely to 

 alteration in the amount of attraction exercised at a given point 

 of the earth's surface. We have but one instrument for estimating 

 atmospheric pressure, that is, the barometer, and it really only 

 gives us the resultant of all the forces, whatever they may be, which 

 act on the atmosphere at the point of observation, without discri- 

 mination of the components. Furthermore, when a barometric 



