VOLCANIC ACTION. 



29 



palaces. I counted thirty greater piles and noted numberless 

 smaller ones ; it was clear that at one time the entire plain 

 had been on fire, that a hundred jets of fire and flame and 

 molten lava had spouted from its surface. The heaps were 

 now old and moss-grown, but we were informed of a vague 

 tradition among the people that their ancestors had seen 

 these flames bursting forth. Altogether, in that important 

 journey, we saw and counted a hundred extinct craters, ex- 

 tending over an arc of ninety miles, not reckoning the central 

 mass of Ank^ratra, round one side of which that arc bends."* 



In a journey to the south-east of Madagascar I discovered 

 traces of volcanic action in many places, in some parts shown 

 in the deposits of rolled pebbles of lava, and in others in the 

 streams of lava rock running into the sea and forming reefs 

 which are being gradually broken up by the surf. And in 

 the very opposite part of the island, on the extreme north- 

 west coast, opposite the Minnow group of islets, Bishop Kestell- 

 Cornish observes : " This coast is the most distinctly volcanic 

 that I have seen in Madagascar ; at one point the lava must 

 have run down to be quenched in the sea, and it looked as if 

 this had taken place only last year." t 



In the Antsihanaka province also the same plutonic agency 

 is distinctly visible. A great part of this region consists of an 

 immense marshy plain, about thirty-five miles long by fifteen 

 wide, with the lake Alaotra at its north-east corner and sur- 

 rounded by hills ; and it has evidently been the seat of some 

 powerful subterranean force by which this depression was 

 caused. This is clear from the fact that the lines of hills 

 which are seen on both sides of the Antsihanaka plain do not 

 run in the same direction as the main valley or depression of 

 the country, but cut it at an angle of about 45°; so that 

 while the general direction of the valley is IST.KE. and S.S.W., 

 the lines of hills on either side have a bearing of N.N.'W. and 

 S.S.E. Many of the ridges seem to be broken off more or 

 less abruptly by the level ground for several miles, and then 

 are continued on the other side of the jolain. It was impos- 

 sible to avoid the conclusion that by some great convulsion 



* Proc. Roy. Geog. Soc, January 25, 1875. 

 t Antananarivo Annual, No. iii. p. 22. 



