the irregularity of this distribution that it is not possible for more than one key to be constructed of the same set of notes. In order to distinguish practically between one and another, one series is taken as the normal key and all the others are severally indicated by expressing the amount of difference between them and it. The normal key, which happens more by accident than design to begin on C, is constructed of what are called Naturals, and all such notes in the entire system as do not occur in this series are called Accidentals. In order to assimilate a series which starts from some other note to the series starting from C, it is necessary to indicate the notes alien to the scale of C, which will have to be substituted for such notes in that scale as could not occur in the new series in other words, to indicate the accidentals which will serve that purpose; and from their number the musician at once recognises the note from which his series must start. This note therefore is called the Key-note, and the artificial series of notes resulting from the arrangement is called the Key. Thus to make a series of notes starting from G relatively the same as those starting from C, the F immediately below G will have to be supplemented by an accidental which will give the necessary semitone between the seventh and eighth degrees of the scale. Similarly, D being relatively the same distance from G that G is from C, the same process will have to be gone through again to assimilate the scale starting from D to that starting from C. So that each time a fifth higher is chosen for a key-note a fresh accidental or sharp has to be added immediately below that note, and the number of sharps can always be told by counting the number of fifths which it is necessary to go through to arrive at that note, beginning from the normal C. Thus C—G, G—D, D—A, A—E is the series of four fifths necessary to be gone through in passing from C to E, and the number of sharps in the key of E is therefore four.
Conversely, if notes be chosen in a descending series of fifths, to present new key-notes it will be necessary to flatten the fourth note of the new key to bring the semitone between the third and fourth degrees; and by adopting a similar process to that given above, the number of flats necessary to assimilate the series for any new key-note can be told by the number of fifths passed through in a descending series from the normal C.
In the Minor Mode the most important and universal characteristic is the occurrence of the semitone between the second and third instead of between the third and fourth degrees of the scale, thereby making the interval between the keynote and the third a minor third instead of a major one, from which peculiarity the term 'minor' arises. In former days it was customary to distinguish the modes from one another by speaking of the key-note as having a greater or lesser third, as in Boyce's Collection of Cathedral Music, where the Services are described as in 'the key of B♭ with the greater third' or in 'the key of D with the lesser third,' and so forth. The modifications of the upper part of the scale which accompany this are so variable that no rule for the distribution of the intervals can be given. The opposite requirements of harmony and melody in relation to voices and instruments will not admit of any definite form being taken as the absolute standard of the minor mode; hence the Signatures, or representative groups of accidentals, which are given for the minor modes are really of the nature of a compromise, and are in each case the same as that of the major scale of the note a minor third above the key-note of the minor scale. Such scales are called relatives—relative major and relative minor—because they contain the greatest number of notes in common. Thus A, the minor third below C, is taken as the normal key of the minor mode, and has no signature; and similarly to the distribution of the major mode into keys, each new key-note which is taken a fifth higher will require a new sharp, and each new key-note a fifth lower will require a new flat. Thus E, the fifth above A, will have the signature of one sharp, corresponding to the key of the major scale of G; and D, the fifth below A, will have one flat, corresponding to the key of the major scale of F, and so on. The new sharp in the former case falls on the supertonic of the new key so as to bring the semitone between the second and third degrees of the scale, and the new flat in the latter case falls on the submediant of the new key so as to bring a semitone between the fifth and sixth degrees. The fact that these signatures for the minor mode are only approximations is however rendered obvious by their failing to provide for the leading note, which is a necessity in modern music, and requires to be expressly marked wherever it occurs, in contradiction to the signature.
There is a very common opinion that the tone and effect of different keys is characteristic, and Beethoven himself has given some confirmation to it by several utterances to the point. Thus in one place he writes 'H moll schwarze Tonart,' i.e. B minor, a black key; and, in speaking about Klopstock, says that he is 'always Maestoso! D♭ major!' In a letter to Thomson of Edinburgh (Feb. 19, 1813), speaking of two national songs sent him to arrange, he says, 'You have written them in , but as that key seemed to me unnatural, and so little consistent with the direction Amoroso that on the contrary it would change it into Barbaresco (qu'au contraire il le cbangerait en Barbaresco), I have set the song in the suitable key.' This is singular, considering his own compositions in the key of four flats, neither of which can justly be entitled barbaresco. Composers certainly seem to have had predilections for particular keys, and to have cast movements in particular styles in special keys. If the system of equal temperament were perfectly carried out, the difference would be less apparent than
- In a sketch for Cello Sonata, op. 102, No. 2, quoted by Nottebohm.
- In a conversation with Rochlitz (Für Freunde der Tonkunst iv. 356).
- Given by Thayer, iii. 45.