It will now be understood how the velarium acted in rendering the solos so clearly audible at the great height and distance of the Upper Press Gallery. Instead of being wasted by diffusion in the great vault above, they were stopped and reflected by the velarium, but not so reflected as to produce disagreeable repetition notes, just audible at particular points, as the lights of the mirror reflections of the chandeliers would be. Flat surfaces reflect radially, while concave surfaces with certain curves reflect sound, light, heat, etc., in parallel lines. The walls and roof of a music-hall should scatter their reflections on all sides, and, therefore, should be flat, or nearly so, excepting at the angles, which should be curved or hollowed. From the orchestra the sound is chiefly required to be projected forward as from the singer’s mouth; and, therefore, an orchestra should have curved walls and roof. Space will not permit a dissertation here on the particular curve required. This has, I believe, been carefully calculated in constructing the Crystal Palace orchestra. Viewed from a distance, the whole orchestra is curiously like a huge wide-opened mouth that only requires to close a little and open a little more, according to the articulations of the choir, to represent the vocal effort of one gigantic throat. There is, I think, one fault in the shape of this mouth. It extends too far laterally in proportion to its perpendicular dimensions. The angles of the mouth are too acute; the choir extends too far on each side. The singers should be packed more like those of the Birmingham Festival Choir. There is an acoustic limit to the magnitude of choirs. Sound travels at about 1100 feet per second, and thus, if one of the singers of a choir is 110 feet nearer than another singer to any particular auditor, the near singer will be heard one-tenth of a second before the more distant, though they actually sing exactly together. In rapid staccato passages this would produce serious confusion, though in such music as most of Handel’s it would be scarcely observable. Some observations which I have made convince me that the actual choir of the Handel Festivals has reached, if not exceeded, the acoustic limits even for Handel’s music, and decidedly exceeds the limits permissible for Mendelsshon and most other composers. I found that when standing on the floor of the building in front of the orchestra, and on one side, I could plainly distinguish the wave of difference of time due to the traveling of the sound, and in all the passages which required to be taken up smartly and simultaneously by the opposite sides of the choir, the effect was very disagreeable.