B. TERRESTRIAL PHYSICS AND METEOROLOGY. 51 



He adduces numerous observations noted by himself, which, 

 in many respects, conflict with the theories that are most 

 commonly held ; and, first, he maintains that the earthquakes 

 of the Caucasus in general can not be attributed directly to 

 volcanic action.. As pictured by him, Shamaka stands at 

 the mouth of the retort of the great natural manufactory of 

 carbureted hydrogen, that in the eastern part of the Cau- 

 casus reveals its products to an, as yet, unknown extent. 

 The chemical actions that are evidently going on within the 

 interior of that portion of the earth afford, therefore, a suffi- 

 cient explanation of the perpetual agitation that the earth 

 experiences. He sees no reason to doubt that Shamaka and 

 its neighborhood will, for the future as in the past, continue 

 to be visited by frequent earthquakes ; but advises the inhab- 

 itants not to remove, since it is possible, by the use of seis- 

 mographs and telegraphic warnings, to so forewarn the peo- 

 ple that no great loss of life need be apprehended. On 

 the other hand, the legalizing of those laws for erecting 

 buildings that seismology, in the hands of Mallet, has long 

 since indicated, would, he thinks, materially diminish the de- 

 struction of public and private buildings. If we understand 

 him correctly, there is but one house in the city that has 

 been built upon the proper principles of safety, and this has, 

 during the past twenty years, outlived every earthquake 

 without the least trace of damage. The question of the 

 proper location of a seismometer is treated by him with some 

 fullness, and he very correctly points out that there are, in 

 fact, but few observatories that enjoy so fortunate a combina- 

 tion of local circumstances as to make them appropriate 

 places for the establishment of these instruments. From the 

 details given by him of the direction of the shock of a number 

 of earthquakes at Tiflis and at Shamaka, we gather that the 

 mean direction of the shocks at Tiflis is from the north-north- 

 west to the south-southeast ; but that very frequent cases of 

 refraction and reflection of the earthquake waves are expe- 

 rienced. Distinguishing between the direction of the shock 

 and the direction of transmission of an earthquake, he finds 

 that the latter occurs always in the direction of the south- 

 west to the northeast. The difference between the direction 

 of the shock and the direction of transmission of an earth- 

 quake is therefore, for that region, so variable and uncertain 



