TRAP PE AN SERIES. 189 



"were poured out. It appears more probable that the 

 latter rocks are really contemporaneous, and that their 

 irregularity is due to their accumulation in the imme- 

 diate vicinity of active vents, — to their having been, in 

 short, part of a volcanic focus. The deep beds of Akub 

 Teriki may be the intrusive mass of the vent itself. 

 The remaining hills, where thick trachyte rests on sand- 

 stone, may be parts of an old lava flow of great depth. 

 Near Fokada there is evidence in favour of a passage 

 from the Senaf^ traps into- those ■ of the neighbourhood 

 of Adigrat. Tt therefore appears most probable that all 

 belong to one group. 



