O'Reilly — De Rossi's Seismical and Endodynmnical Map. 457 



observation is made, it is always assumed that tlie atmosphere is in 

 a normal state as regards its constitution, at least for those alti- 

 tudes of which we have some knowledg-e. But that is simply 

 assumption, and the constitution may he abnormal at a given time 

 and for a given locality ; and that such frequently occurs imme- 

 diately before earthquakes may be gathered from some of the facts 

 recorded as having preceded many earthquakes, and which have 

 not, perhaps, received the full amount of attention which they 

 merit. Thus there is frequently made mention of indications of 

 uneasiness, more or less marked, on the part of certain classes of 

 animals, more particularly of those which, by reason of their size 

 or their habits, are naturally brought into close contact with the 

 earth, such as small quadrupeds, fowl, cats, dogs, snakes, &c. Now 

 this may fairly and reasonably be attributed to the emission of 

 gases of a nature hurtful to such animal life, such as carbonic acid, 

 certain carbides of hydrogen, hydrogen itself, or even sulphuretted 

 hydrogen, all which are intimately associated with volcanic and 

 seismic manifestations. This view is strengthened by the frequent 

 alterations reported as occurring before earthquakes in springs and 

 streams, such as their drying up or disappearing by reason of joints 

 or fissures presenting an easy escape for gases which may be im- 

 prisoned in depths, and the excessive pressure of which is a main 

 cause of many if not of most earthquakes. How long before the 

 earthquake may such escapes of gas have been going on on a minute 

 scale, hardly, if at all noticeable, is a question of much interest, and 

 would point to the importance to be attached to the careful exami- 

 nation of the waters of springs, particularly if they have their 

 sources in districts at all affected by earthquakes. The well-known 

 variations in the volume and temperature of already existing hot- 

 springs consequent on earthquakes justify this point of view as 

 well as the equally well-established role of these in giving rise to 

 such springs. We can thus see that there is an intimate relation 

 to be recognised between earthquake action and emission of gases, 

 and between these and Assuring or jointing. Now, the tendency of 

 all recent seismical research and observation is to connect earth- 

 quake action with the existence of systems of jointing and fissur- 

 ing. And thus we have a relation established between these 

 emissions of gases and earthquakes. But such gases, by reason of 

 their specific densities and temperatures of emission, influence more 



