We hastened from this sickly spot, trying to take the attentions of the mosquitoes as hints to seek more pleasant quarters on the healthy shores of Lake Nyassa; and when we sailed into it, on the 2nd September, we felt refreshed by the greater coolness of the air off this large body of water.  The depth was the first point of interest.  This is indicated by the colour of the water, which, on a belt along the shore, varying from a quarter to half a mile in breadth, is light green, and this is met by the deep blue or indigo tint of the Indian Ocean, which is the colour of the great body of Nyassa.  We found the Upper Shire from nine to fifteen feet in depth; but skirting the western side of the lake about a mile from the shore the water deepened from nine to fifteen fathoms; then, as we rounded the grand mountainous promontory, which we named Cape Maclear, after our excellent friend the Astronomer Royal at the Cape of Good Hope, we could get no bottom with our lead-line of thirty-five fathoms.  We pulled along the western shore, which was a succession of bays, and found that where the bottom was sandy near the beach, and to a mile out, the depth varied from six to fourteen fathoms.  In a rocky bay about latitude 11 degrees 40 minutes we had soundings at 100 fathoms, though outside the same bay we found none with a fishing-line of 116 fathoms; but this cast was unsatisfactory, as the line broke in coming up.  According to our present knowledge, a ship could anchor only near the shore.

Looking back to the southern end of Lake Nyassa, the arm from which the Shire flows was found to be about thirty miles long and from ten to twelve broad.  Rounding Cape Maclear, and looking to the south-west, we have another arm, which stretches some eighteen miles southward, and is from six to twelve miles in breadth.  These arms give the southern end a forked appearance, and with the help of a little imagination it may be likened to the “boot-shape” of Italy.  The narrowest part is about the ankle, eighteen or twenty miles.  From this it widens to the north, and in the upper third or fourth it is fifty or sixty miles broad.  The length is over 200 miles.  The direction in which it lies is as near as possible due north and south.  Nothing of the great bend to the west, shown in all the previous maps, could be detected by either compass or chronometer, and the watch we used was an excellent one.  The season of the year was very unfavourable.  The “smokes” filled the air with an impenetrable haze, and the equinoctial gales made it impossible for us to cross to the eastern side.  When we caught a glimpse of the sun rising from behind the mountains to the east, we made sketches and bearings of them at different latitudes, which enabled us to secure approximate measurements of the width.  These agreed with the times taken by the natives at the different crossing-places—­as Tsenga and Molamba.  About the beginning of the upper third the lake is crossed by taking advantage of the island Chizumara, which name in the native tongue means the “ending;” further north they go round the end instead, though that takes several days.