In the intermediate ages we sometimes find works of this sort attributed to the Bishop alone, sometimes to the Chapter alone, sometimes to the Bishop and Chapter working together. I suspect that this last would commonly be the truer account in all cases; at any rate, what either Bishop or Chapter did the other party must have consented to. Jocelin was doubtless the great mover in the work, the life and soul of the whole undertaking. The whole would be done under his care, and his personal contributions would doubtless be large. But all this in no way shuts out the co-operation of the Chapter, of the clergy and laity of the diocese, and of well-disposed persons wherever they might be found.[39] Another part of the buildings of the church belongs to the age of Jocelin, where his hand might not have been looked for at first. This is the cloister as it stood in its first estate. You will remember that the cloister which was built by Gisa, together with his dormitory and refectory, was pulled down by John of Tours. You will also remember that a cloister, which in a monastery is an essential part of the building, and is always built after a particular model, is in a secular church a mere convenience, which may perfectly well be left out, and which may be built in any place and after any fashion which may be thought good. Jocelin then, or his Canons, now built them a cloister, but it was a cloister which was no longer accompanied, as in Gisa's time, by any refectory or dormitory. It is more like a monastic cloister than those of Chichester and Hereford; it is less like one than that of Salisbury. It occupies, like a monastic cloister, one side of the nave; still it is not a perfect square, but an irregular parallelogram; it has no walk on the north side, and the eastern walk comes up against the south end of the transept, while in a monastery it would have been built against its western wall. To the east, where the chapter-house would have stood in a monastery, there was a detached Lady chapel, of which the traces may easily be seen, but which was rebuilt late in the fifteenth century and wholly destroyed in the sixteenth. Now that the cloister was first built at this time is plain, as all the outer walls, including that very pretty doorway leading to the Palace, are all of Jocelin's date. The doorway leading from the transept into the cloister is also mentioned in an Act of Chapter in 1297, printed in Dugdale's Monasticon.[40] But this very doorway, and the doorway which is in some sort the fellow to it in the south-western tower, give us the surest signs that the cloister is not now in the same state in which it was originally designed. Even in its first estate, it seems to have been, as we should expect, an addition, though an addition made not very long after the building of the part of the church which it joins. The wall comes up uncomfortably close against this fine doorway, though it does not mutilate it in the way