we have 11 deg. 40’ as the latitude of Manda.  The mainland to the south of Mpabala is called Kabende.  The land’s end running south of Masantu’s village is the entrance to the Luapula:  the clearest eye cannot see across it there.  I saw clouds as if of grass burning, but they were probably “Kungu,” an edible insect, whose masses have exactly the same appearance as they float above and on the water.  From the time the canoes take to go to Kabende I believe the southern shore to be a little into 12 deg. of south latitude:  the length, as inferred from canoes taking ten days to go from Mpabala to the Chambeze, I take to be 150 miles, probably more.  No one gave a shorter time than that.  The Luapula is an arm of the Lake for some twenty miles, and beyond that is never narrower than from 180 to 200 yards, generally much broader, and may be compared with the Thames at London Bridge:  I think that I am considerably within the mark in setting down Bangweolo as 150 miles long by 80 broad.

When told that it contained four large islands, I imagined that these would considerably diminish the watery acreage of the whole, as is said to be the case with five islands in Ukerewe; but even the largest island, Chirubi, does not in the least dwarf the enormous mass of the water of Bangweolo.  A range of mountains, named Lokinga, extends from the south-east to the south-west:  some small burns come down from them, but no river; this range joins the Kone, or Mokone range, west of Katanga, from which on one side rises the Lufira, and on the other the Liambai, or Zambesi.  The river of Manda, called Matanga, is only a departing and re-entering branch of the Lake, also the Luma and Loela rivers—­some thirty yards broad—­have each to be examined as springs on the south of the Lake.

July 29th, 1868.—­Not a single case of Derbyshire neck, or of Elephantiasis, was observed anywhere near the Lake, consequently the report we had of its extreme unhealthiness was erroneous:  no muddy banks did we see, but in the way to it we had to cross so many sponges, or oozes, that the word matope, mud, was quite applicable; and I suspect, if we had come earlier, that we should have experienced great difficulty in getting to the Lake at all.

30th July, 1868.—­We commenced our march back, being eager to get to Chikumbi’s in case Mohamad should go thence to Katanga.  We touched at Mapuni’s, and then went on to the Molongosi.  Clouds now began to cover the sky to the Mpanda, which has fifteen yards of flood, though the stream itself is only five yards wide, then on to the Mato and Moiegge’s stockade, where we heard of Chikumbi’s attack on Kombokombo’s.  Moiegge had taken the hint, and was finishing a second line of defence around his village:  we reached him on the 1st August, 1868, and stopped for Sunday the 2nd:  on the 3rd back to the Rofubu, where I was fortunate enough to hire a canoe to take me over.