A DEEP LAKE 199 



and three to four feet beam, hollowed out of a single tree. 

 We all embarked and got off soon after half-past seven, but the 

 wind had already risen somewhat, and there was quite a swell 

 on the water. But the sail across was most delightful. As we 

 proceeded, the northern shores opened up, showing two deep 

 bays stretching far away between the hills, and an island, where 

 the Sihanaka made their last stand hi resisting Hova domina- 

 tion. From that time it has not been allowed to be inhabited, 

 but is only used for planting vegetables. We had only two 

 paddlers, one at the head, and the other at the stern of the 

 canoe, and so were an hour and ten minutes in crossing. We 

 made an attempt to ascertain the depth of the lake with an old 

 knife as a sinker, and a piece of string as a line, while the doctor, 

 in true scientific fashion, " hove the lead." I regret to say that 

 no accurate information was obtained, for the sounding line was 

 again and again thrown with the report, " no bottom." But our 

 short line was no doubt the reason of our ill-success. The lake 

 is probably deep at its northern end, and it is certainly shallow 

 at its southern extremity, gradually changing into marsh. Some 

 of my missionary friends, who subsequently lived in Antsi- 

 hanaka, have described voyages across the southern end of the 

 Alaotra, where, amongst the dense growth of papyrus, rush, and 

 tall grasses, the only practicable paths for a canoe are dark 

 passages, almost tunnel-like, among the rank vegetation ; and 

 where a stranger might easily be lost in the watery and reedy 

 wastes around him. 



There can be no doubt that the present lake is but a small 

 remnant of a much larger one ; for, at a not very distant 

 period, the water must have covered the whole plain of Antsi- 

 hanaka, thus forming a lake five or six tunes the size of the 

 present Alaotra. But at a yet earlier period still, this lake 

 extended for a hundred miles farther south, down the Ankay 

 plain, and for at least two hundred miles farther north, 

 forming an immense extent of water, not much unlike the 

 Tanganyika in Central Africa in size and outline, and of con- 

 siderable depth ; for Mr Baron found numerous indications of 

 old shore-lines at elevations of eleven to twelve hundred feet 

 above the present level. Doubtless, the gradual lowering of 

 the valleys of the Mangoro to the south, and of the Maning6ry 

 to the north-east, drained off this great lake, leaving only 



