1st. Columnar spokes and no exterior openings, as at St. Stephen’s at Beauvais, and the window over the choir at Lyon.
2nd. Columnar spokes and exterior openings. In France we find such at Chartres, and in the end windows of the transept of which I am now treating: to both these I should give the name of wheel windows.
3rd. No columns; the divisions are variously branched, but still exhibiting an appearance of radiation. Such as this we have at Amiens, Beauvais, Lyons, and many other places, and I should appropriate to them the name of rose windows.
4th. Arches and ornaments arising from the circumference as well as from the centre. This disposition gives a squareness to the ends of the divisions which may well merit the name of marigold windows, the cathedrals at Mantes and at Chalons sur Marne, will offer examples.
5th. No radiation preserved in the principal divisions. I do not know that I can cite for this any other example than that at the cathedral at Troyes, and it may, without inconvenience, be left without a name.
I believe these different arrangements succeeded each other nearly in the order I have mentioned, but not uniformly so. In small windows of the same epoch, the disposition is generally more simple than in the large ones, but after the columnar spokes had once been abandoned, it does not appear that they were ever resumed.
The gallery or triforium, of the transept and choir, has semi-circular arches resting on columns almost Corinthian, and on pilasters which might be deemed of the _renaissance_, if some of the latter were not zig-zag. On the whole, if I had met in the north of France with a building corresponding in character with the choir and transept of the cathedral at Lyon, I should say that it had been erected about the year 1200, or rather earlier, when the first style of pointed architecture was beginning to give way to the second. In the nave, the larger shafts are connected with the masonry of the piers and walls, the smaller are constructed separately. At Nôtre Dame at Dijon, exactly the reverse takes place; there, the large shafts have an independent construction, and the smaller are united with the mass of the work. Some of the pillars next the choir, as well as those of the choir itself, have nearly the ancient Attic base; in others, the Gothic forms are fully developed. The groining of the vaults is oblique, and the last pair of pillars seems to be an addition of the fourteenth century. In the fifteenth century, several chapels were added to the nave; the last and most beautiful of which is that which was built for Charles, Cardinal de Bourbon, who was king of France for four hours. This Charles, Duke de Vendôme, Cardinal Archbishop of Rouen, and legate of Avignon, was born in 1523, put upon the throne in 1589 by the Duke de Mayenne, and died in 1590; is it possible we can have Gothic architecture in this city of so late a date? My guide-book tells me that his brother, Pierre de Bourbon, who finished this chapel, married the daughter of Louis XI. and multiplied the thistle among his ornaments, to signify that the king had made him a ‘_cher don_.’ Louis XI. died in 1483, and I suspect that my history is not correct. This chapel is entirely in the pointed style, and part of the vaulting exhibits some indication of the manner of our Henry the Seventh’s chapel at Westminster. We find also the bases of adjoining parts on different levels, and mouldings lost and re-appearing, or seeming to pass one behind the other; but I cannot find the complicated arch, so common in the late French Gothic, in any of the ornaments.