[17] This grain was probably what is now well known under the name of
    millet.—­E.

[18] According to Barros, Aguada da boa Paz.—­Clarke.

[19] Gibb’s Orosius, I. 50.

[20] The text here ought probably to be thus amended, “He and his brother,
    with Nicholas Coelle,” &c.—­E.

[21] These probably swam off to the ships.—­E.

[22] De Faria alleges that the people of this river were not so black as
    the other Africans, and wore habits of different kinds of stuffs, both
    cotton and silk, of various colours, and that they understood Arabic;
    and adds, that they informed De Gama there were white people to the
    eastwards, who sailed in ships like those of the Portuguese.  Osorius
    likewise says, that one of the natives spoke Arabic very imperfectly,
    and that De Gama left two of his convicts at this place, which he
    called San Rafael.—­Clarke.

[23] There is no circumstance in the text from which the situation of this
    river can even be conjectured.  Clarke, p.440, alleges that it was
    Soffala; and yet, in a note in his preceding page, says, “That De Gama
    seems to have passed Cape Corientes during the night, and to have kept
    so far from land, on account of a strong current setting on shore, as
    not to have noticed Sofala.”  In the notes on the Lusiad, this river of
    Good Signs is ascertained to have been one of the mouths of the
    Zambeze, or Cuama River, which divides Mocaranga from the coast of
    Mozambique; the different mouths of which run into the sea between the
    latitudes of 19 deg. and 18 deg.  S.—­E.

[24] They were evidently afflicted with the scurvy; and accordingly De
    Barros refers the disease to its proper cause, “Having been for so
    long a time confined to the use of salt fish and corrupted biscuit.—­
    Clarke.”

[25] Addition to the narrative of Castaneda, from De Barros.—­Clarke.

[26] This obscure expression seems to mean that De Gama wished them to
    precede the ships, and point out the way into the harbour.—­E.

[27] This expression has probably been misunderstood by the original
    translator.  It appears that these Moors of Mozambique spoke Arabic,
    here called the language of Algarve, and finding themselves understood
    and answered by the strangers, mistook the Portuguese for Moors.—­E.

[28] Mozambique is in lat. 15 deg. 35’ S. and in 41 deg. of E. Long—­E.

[29] The observations here inserted, and marked with inverted commas, are
    made by the Editor of the present collection.  They are much too long
    for insertion in the form of a note, and appeared of too much
    importance to be omitted; being chiefly from Clarke, I. 447.—­E.

[30] For the materials of this addition to the text of Castaneda, we are
    chiefly indebted to the Progress of Maritime Discovery, p. 447, 458. 
    —­E.