The dagala or nsipe, a small fish caught in great numbers in every flowing water, and very like whitebait, is said to emit its eggs by the mouth, and these immediately burst and the young fish manages for itself.  The dagala never becomes larger than two or three inches in length.  Some, putrefied, are bitter, as if the bile were in them in a good quantity.  I have eaten them in Lunda of a pungent bitter taste, probably arising from the food on which the fish feeds.  Men say that they have seen the eggs kept in the sides of the mouth till ready to go off as independent fishes.  The nghede-dege, a species of perch, and another, the ndusi, are said to do the same.  The Arabs imagine that fish in general fall from the skies, but they except the shark, because they can see the young when it is cut open.

10th July, 1869.—­After a great deal of delay and trouble about a canoe, we got one from Habee for ten dotis or forty yards of calico, and a doti or four yards to each of nine paddlers to bring the vessel back.  Thani and Zahor blamed me for not taking their canoes for nothing; but they took good care not to give them, but made vague offers, which meant, “We want much higher pay for our dhows than Arabs generally get:”  they showed such an intention to fleece me that I was glad to get out of their power, and save the few goods I had.  I went a few miles, when two strangers I had allowed to embark (from being under obligations to their masters), worked against each other:  so I had to let one land, and but for his master would have dismissed the other:  I had to send an apology to the landed man’s master for politeness’ sake.

[It is necessary to say a few words here, so unostentatiously does Livingstone introduce this new series of explorations to the reader.  The Manyuema country, for which he set out on the 12th of July, 1869, was hitherto unknown.  As we follow him we shall see that in almost every respect both the face of the country and the people differ from other regions lying nearer to the East Coast.  It appears that the Arabs had an inkling of the vast quantities of ivory which might be procured there, and Livingstone went into the new field with the foremost of those hordes of Ujijian traders who, in all probability, will eventually destroy tribe after tribe by slave-trading and pillage, as they have done in so many other regions.]

Off at 6 A.M., and passed the mouth of the Luishe, in Kibwe Bay; 3-1/2 hours took us to Rombola or Lombola, where all the building wood of Ujiji is cut.

12th July, 1869.—­Left at 1.30 A.M., and pulled 7-1/2 hours to the left bank of the Malagarasi River.  We cannot go by day, because about 11 A.M. a south-west wind commences to blow, which the heavy canoes cannot face; it often begins earlier or later, according to the phases of the moon.  An east wind blows from sunrise till 10 or 11 A.M., and the south-west begins.  The Malagarasi is of considerable size at its confluence, and has a large islet covered with eschinomena, or pith hat material, growing in its way.