These changes, you will see, finished the ground-plan of the church itself as it now stands. The church itself has not been extended northwards, southwards, eastwards, or westwards, since the days of Bishop Ralph. Nor, on the other hand, has any part of the church itself been destroyed. Other buildings have been attached to it, and parts of the subordinate buildings have perished, but the ground covered by the church itself is exactly the same now as it was when Ralph was buried before the high altar. As a church then, as a building set apart for divine worship, Saint Andrew's was now quite perfect and needed neither addition nor change. Nave, choir, presbytery, chapels, and the one necessary adjunct of the Chapter-house, were all finished. But besides the completion of the ground-plan, there was another great work to be done before the building could be said to be finished as a work of architecture. Jocelin had not carried his three towers above the height of the roofs; they were mere stumps, and the effect must have been unfinished and unsightly. In the course of the fourteenth and fifteenth centuries this defect was supplied. Indeed, as far as the central tower is concerned, the defect had been supplied already. I have carried on the history of the changes which affected the ground-plan as a continuous narrative, but the raising of the central tower and its consequences belong to the same period. The raising of the tower seems to have formed part of the general plan of recasting the whole part of the church east of the crossing, and it may actually have been the first instalment of the work. I may here perhaps say a few words on the general subject of central towers. As the principal feature of churches of the highest class, the central tower is all but confined to England and Normandy; in other parts of France it is common enough, but, reversing our English rule, it is common in churches of a smaller class, but nearly unknown in the great cathedrals and abbeys. I ought perhaps to say that I am now speaking mainly of Gothic buildings, not of Romanesque. The truest way of putting the case would perhaps be that the central tower, the direct representative of the cupola, is a Romanesque feature, prevalent everywhere in Romanesque times, but which England and Normandy alone retained in large churches of later date. The question of central tower or no central tower resolves itself into this; which is the greater merit in a cathedral or other great church--the highest amount of internal majesty, or the highest perfection of external outline? England and Normandy decided for the external outline; the rest of Western Christendom decided for the internal effect. A great French church, Amiens, Beauvais, Chartres, Rheims, Saint Quentin, is carried up to a height in the inside of which we in England have no notion. But this internal majesty is bought by the utter sacrifice of external outline. The crossing of the four