Vol. 51.] OP A JOURNEY IN MADAGASCAR. 61 



of the southern half) is the preponderance, among the rocks exposed, 

 of dolerite. From Fenoarivo to, say, lat. 14° S. 1 — a distance of 

 about 200 miles (and my journey led me for the most part along 

 the shore) — one sees comparatively little else than this dolerite ; 

 occasionally, however, the underlying gneiss comes to the surface. 

 The question arises, therefore, — 'Whence has it been derived ? The 

 only satisfactory explanation of its occurrence seems to be that 

 it has flowed from large fissures farther inland, which at present 

 are represented by the dykes above referred to. This I conclude 

 from the following facts : — (1) The exact similarity, for the most 

 part, in mineral composition and texture of the dyke-rock with that 

 on the coast. I have examined microscopically several dozen 

 sections of both. They consist almost always essentially of plagio- 

 clase and augite, and show subophitic, or very seldom typically 

 ophitic, texture. Occasionally they contain also a slight amount of 

 olivine. (2) The persistence of this one type of rock for so great a 

 distance along the coast in a north-and-south direction. This is 

 easily accounted for on the supposition of a dyke or series of dykes 

 a few miles inland (which, as before stated, actually exist) running 

 more or less parallel with the coast, from which lava has issued. 

 (3) In some places the rock is distinctly traceable as a lava-sheet 

 for several miles inland ; and as there is no large volcanic vent or 

 series of vents ranging parallel with the coast from which it could 

 have flowed, the dykes above referred to must be looked upon as its 

 source. For these reasons I conclude that fissure-eruption has 

 occurred on a large scale on the eastern side of Madagascar. The 

 subophitic and ophitic textures also seem to suggesjmjhat the flow has 

 been of considerable thickness. 



The lava-sheet has been, at least in some places, enormously thick. 

 A few miles north of the coast-town of Antalaha (lat. 14° 57' S.), for 

 instance, it covers the country to a great depth for several miles 

 inland — how many I cannot say. In other localities numerous hills, 

 some hundreds of feet in height, composed entirely of dolerite, stand 

 as witnesses to the great volume of the lava-flow. This thickness 

 may account for the prevalently doleritic texture of the rock ; only 

 occasionally is it basaltic. The lava has in many places been 

 entirely removed by denudation, and in other places it is covered 

 by dense, impenetrable forest. The rock is almost always compact 

 in texture, though in a few places it is vesicular : for example, at 

 the locality known as Ambatofaingainy (532), lat. 15° 8' S., about 

 halfway between ISTgontsy and Antalaha, where it is amygdaloidal, 

 the vesicles being filled in with calcite, quartz, dark green chal- 

 cedony, chlorite, and zeolites ; the cracks in the rock being occupied 

 by ferruginous matter and chalcedony. The rock at this locality is a 

 basalt, or perhaps an andesitic basalt. A quite exceptional type of 

 dolerite occurs at Antalaha itself (27 J), the tolerably large crystals 

 of plagioclase and augite being embedded in brown tachylytic glass. 



1 When the longitude is not given, the place indicated is on or near the sea- 

 coast. 



