Thus even in Saxon times the choir of Canterbury, which extended into the nave, was enclosed by a breast-high wall. And Gervase tells us of the choir of Conrad (A.D. 1130), “that at the bases of the pillars there was a wall built of marble slabs, which, surrounding the choir and presbytery, divided the body of the church from its sides called aisles.” The same choir of Canterbury is now enclosed by the very beautiful screen built by Prior d’Estria. Rochester again has a solid stone wall round its choir, and the same is the case at _Lincoln_ (though never a monastery). “The choir, as a rule, was occupied,” says Dr. Jessopp, “exclusively by the monks or nuns of the monastery. The servants, work-people, and casual visitors who came to worship were not admitted into the choir; _they_ were supposed to be present only on sufferance. The church was built for the use of the monks: it was their private place of worship.” And, as we shall see presently, the screens were still more solid where there happened to be a parish church in the nave. Pugin, in his well-known work on the subject of rood-screens, points out another reason for these walled-in choirs (besides the one just mentioned), _i.e._ that they shut off in some degree the cold draughts of air from the monks during their frequent and lengthy services in buildings which at that time (and for many a century to come) were not warmed at all. Later still the cathedrals, other than monastic, like _Lincoln_, and collegiate churches like Southwell, followed suit in separating off the choir from the nave. In all these cases, as a rule, the screen was solid, but when the same movement spread to parish churches, the chancel was divided off by a screen of open work, there being no need for the N. and S. screens, as a rule (though Newark has them), so that the congregation—here regarded as an integral part of the worshippers—might see the altar and all the ceremonial. In such Saxon churches as have retained their Saxon chancel arches, these are very narrow, and originally in all probability were closed by a veil, more or less completely. This is alluded to in an Anglo-Saxon Pontifical, “extenso velo inter eos (_i.e._ clericos) et populum,” and by Durandus (who was Bishop of Mende, a small city in the Lozère district of France from 1286 to 1296), “interponatur velum aut murus inter clerum et populum.” Later, when the use of the veil ceased, and these chancel arches were felt to be, through their narrowness, a great obstruction to the view from the nave into the choir, openings were made on each side to partially circumvent this difficulty. At _Bracebridge_, near Lincoln, for example, the chancel arch, exceedingly narrow (being only 5 feet wide), is of Saxon date and has a round-headed opening—a hagioscope—on either side.