There seems to be some uncertainty here as to what Cecil Sharp was talking about when he referred to “the song-books of the past”. These would have been collections of early 19th century art music compositions, certainly including the works of Charles Dibdin - whose songs Sharp made a point of saying had almost completely died out amongst the peasantry in spite of their previous popularity - and Henry Bishop, whose ‘Home Sweet Home’ he likewise never heard in the field. He might also have been thinking of the British Songster of 1786, which contained the hits of the day from the pleasure gardens, though I haven’t found any reference to it in his writings. His point was that composed ‘art-songs’ of this kind were not only qualitatively different from the songs he described as ‘folk’, but had no been adopted by ‘the folk’. He gave due consideration to the idea that folk songs might be ‘garden escapes’ from the popular or art music of the past, but finally came down on the side of the theory that they were “the musical creations of the common people”, albeit through the cumulative effects of variation and selection rather than necessarily in their original composition.