94 REPORTS ON THE STATE OF SCIENCE. 



Derbent. 



The town lies on the lower shore of the Caspian Sea. 



The nearest point of interest from the town is the mountain Dschalgan, whose 

 steep brows approach the town from the south-west. To the southerly side of the 

 mountain we find a synclinal fault which stretches N.W.-S.E., and dips south-east. 

 In the direction of the town the tectonic structure becomes tolerably complicated, 

 and the connection of tlie Dschalgan synclinal with the flat dome-shaped anticlinal 

 which comes to sight on the north side of the town is not yet explained. Tertiary 

 deposits as well as the Quaternary Loess formation are also found. The latter 

 deposit stretches from the sea-shore at Derbent to a height of 100 fathoms up the 

 brow of the Dschalgan. 



In the sea-clifEs, not far from the railway station, shell lime rocks come to view. 



The remainder of the Dschalgan heights are formed of limestone, sandstone and 

 clay. Along the south-east brows of the Dschalgan, not far from the fortress, one 

 finds the following strata cropping up in ascending order :— 



1. Grey and cofEee-brown coloured sandy clay. 



2. 3-2 inches of clay strata with ferruginous grains. 



3. Layers of white sandy clays. 



4. Grey shell limestone (holding Mactra), 17-15 fathoms thick. 



5. Dividing layers of marl clay with broken pieces of Mactra shells, 3 inches. 



6. Grey Mactra shell limestone, 14-15 fathoms. 



7. Dividing layer clay marl, 14 inches. 



8. Yellowish solid limestone with few large shells of Mactra, such as Vitaliana, 

 d'Ori). 



Tiftis. 



The town lies on the slopes of the valley of the river Kura. 



With the exception of a few places, where primitive rock comes to view, the 

 surface formation of the Tiflis valley is composed of clayey sandstone and loam 

 deposits, whose thickness amounts in places to 80 feet and more. This is covered with 

 horizontal strata of alluvium. 



The heio'hts in the environs of the town consist of the above rocks, together with 

 sedimentary and volcanic rock debris, which are cemented together by a clayey 



medium. 



In an exposure on the right bank of the Kura, in the immediate neighbourhood 

 of the seismic station, one finds more or less thinly laminated clay sandstone of a 

 grey colour ; the same stratification is to be seen in the bottom of a deep trench 

 which surrounds the station. The foundations of the station, as well as all 

 seismographs, are directly set on the original rock. 



III. The Origins of Large Earthquakes in 1 905. 



In registers from dififerent stations in 1905, as in other years, the 

 number of entries vary within wide limits. In the list for Shide there 

 are 159 entries, but 47 of these refer to extremely minute displacements 

 the nature of which is uncertain. Disturbances which are undoubtedly 

 of seismic origin are therefore 112. Out of these 56 were distinctly 

 megaseismic. On the accompanying map (Plate I.) the origins of 57 

 widespread movements are indicated. This number happens to be the 

 annual average for the years 1899 to 1905 inclusive. 



A glance at the map shows, with but few exceptions, that these earth- 

 quakes are confined to a circle passing from Central America through the 

 Azores, the Alpine, Balkan and Himalayan ranges, into the East Indian 

 Archipelago. 



The quiescence of districts not lying on this band is very marked. 



Destructive earthquakes occurred on April 4 in N.W. India, and on 

 September 8 in Calabria. Whether the latter was in any way connected 

 with the relief of volcanic stress which commenced in May 1905 and 



