Some difficulty was experienced by those who have transcribed the music of the negroes in reproducing 'the entire character' of the songs by the conventional symbols of the art. This is due in part to the primitive elements in the music, and in part to the peculiar manner of the performance. The characteristic improvisational style of the negro, the peculiar quality of the voices, and the slurring of certain values are all necessary in order to produce the proper effect. Moreover, the improvised harmony, simple as it was, had become an inherent part of the music not easily to be reproduced. The following description, taken from 'Slave Songs in the United States,' may be illuminating in this connection: 'There is no singing in _parts_, as we understand it, and yet no two appear to be singing the same thing; the leading singer starts the words of each verse, often improvising, and the others who "base" him, as it is called, strike in with the refrain, or even join in the solo when the words are familiar. When the "base" begins the leader often stops, leaving the rest of the words to be guessed at, or it may be they are taken up by one of the other singers. And the "basers" themselves seem to follow their own whims, beginning when they please and leaving off when they please, striking an octave above or below (in case they have pitched the tune too high), or hitting some other note that "chords," so as to produce the effect of a marvellous complication and variety and yet with the most perfect time and variety, and yet rarely with any discord. And what makes it all the harder to unravel a thread of melody out of this strange network is that, like birds, they seem not infrequently to strike sounds that cannot be precisely represented by the gamut and abound in "slides" from one note to another and turns and cadences not in articulated notes.' A word should be added here regarding the instruments used by the negro. The one most closely identified with him is, of course, the banjo, which, in a primitive form, he is said to have brought from Africa. The 'banjar' to which Thomas Jefferson refers in his 'Notes on Virginia' was an instrument of four strings, or perhaps less at first, whose head was covered with a rattlesnake's skin, and which resembled closely an instrument used by the Chinese. (_Cf._ Vol. I, p. 54.) It is thought that the original banjo was a melodic rather than a harmonic instrument, which is the peculiar office of its modern off-spring, and, since the negro's music was at first purely melodic, it must have been accordingly played. The tuning, too, was probably very different from that of the banjo of to-day.