SUJ^FACE AND GEOLOGICAL STRUCTURE 
19 
the primitive formation of the region of high land; the 
earlier subterranean activity was obviously widely spread. 
It has already been stated that numerous volcanic cones 
occur in the neighbourhood of the Itasy lake, and accord¬ 
ing to Dr. Mullens, about a hundred extinct volcanoes 
can be observed in that region, within a circuit of 96 
miles. In some places the lava looks as fresh, as black 
and as sharp as if it had but just burst out from the 
crater. Though the volcanoes are not at present active, 
there is a tradition amongr the natives that their ancestors 
had seen great flames there in olden times. The lofty 
Ankaratra mountains are also extinct volcanoes. The 
Rev. Mr. Campbell remarks that large quantities of scoriae 
lie scattered around and convey the impression that the 
region has been a great smelting-house. Lava detritus 
also occurs in the south-east, and broad streams of lava 
extend as far as the sea-coast. 
Extinct volcanoes are also mentioned in the north, in 
the land of the Sihanaka. According to the statement 
of Kestell-Cornish, several lava masses of quite recent 
appearance descend to the sea in the north-west, and 
the volcanic soil is known to extend beyond Nossi-Be 
to the Comoro isles. 
While later eruptive rocks and primitive mountain 
masses of great antiquity lie upon the upper part of 
the pedestal of granite, there are genuine sedimentary 
rocks lower down. These are mesozoic as well as 
early tertiary strata of marine origin (eocene). Traces 
of these, however, only exist on the west side, while 
on the eastern side formations of this kind appear to 
be entirely wanting. Vinson indeed asserts that he has 
seen sedimentary strata on this side, full of fossil remains, 
but others have been unable to discover anything of 
the kind. 
In Betsimisaraka I have only been able to observe 
mighty layers of “losz” along the rivers and in the few 
