21st August, 1866.—­Started for the Loangwa, on the east side of the Lake; hilly all the way, about seven miles.  This river may be twenty yards wide near its confluence; the Misinje is double that:  each has accumulated a promontory of deposit and enters the Lake near its apex.  We got a house from a Waiyau man on a bank about forty feet above the level of Nyassa, but I could not sleep for the manoeuvres of a crowd of the minute ants which infested it.  They chirrup distinctly; they would not allow the men to sleep either, though all were pretty tired by the rough road up.

22nd August, 1866.—­We removed to the south side of the Loangwa, where there are none of these little pests.

23rd August, 1866.—­Proposed to the Waiyau headman to send a canoe over to call Jumbe, as I did not believe in the assertions of the half-caste Arab here that he had sent for his.  All the Waiyau had helped me, and why not he?  He was pleased with this, but advised waiting till a man sent to Losewa should return.

24th August, 1866.—­A leopard took a dog out of a house next to ours; he had bitten a man before, but not mortally. 29th August, 1866.—­News come that the two dhows have come over to Losewa (Losefa).  The Mazitu had chased Jumbe up the hills:  had they said, on to an island, I might have believed them.

30th August,1866.—­The fear which the English have inspired in the Arab slave-traders is rather inconvenient.  All flee from me as if I had the plague, and I cannot in consequence transmit letters to the coast, or get across the Lake.  They seem to think that if I get into a dhow I will be sure to burn it.  As the two dhows on the Lake are used for nothing else but the slave-trade, their owners have no hope of my allowing them to escape, so after we have listened to various lies as excuses, we resolve to go southwards, and cross at the point of departure of the Shire from the Lake.  I took lunars several times on both sides of the moon, and have written a despatch for Lord Clarendon, besides a number of private letters.

3rd September, 1866.—­Went down to confluence of the Misinje and came to many of the eatable insect “kungu,”—­they are caught by a quick motion of the hand holding a basket.  We got a cake of these same insects further down; they make a buzz like a swarm of bees, and are probably the perfect state of some Lake insect.

I observed two beaches of the Lake:  one about fifteen feet above the present high-water mark, and the other about forty above that; but between the two the process of disintegration, which results from the sudden cold and heat in these regions, has gone on so much that seldom is a well-rounded smoothed one seen; the lower beach is very well marked.

The strike of large masses of foliated gneiss is parallel with the major axis of the Lake, and all are tilted on edge.  Some are a little inclined to the Lake, as if dipping to it westwards, but others are as much inclined the opposite way, or twisted.