and its aisles are but seventy-four feet wide, we infer that the additional twenty-three feet, of the crypt's width, must underlie the vestries. This southern adjunct of the crypt, beneath the vestries, is divided into two alleys, by a row of four columns. A solid wall separates it from the south aisle of the actual crypt. The great crypt, seventy-four feet wide, is divided into eight alleys; there being four in the central space, and two in each aisle. Each rank of columns contains ten of them, in the total length of seventy feet, and the shafts are, consequently, more slender than those usually found in cathedral crypts. As there are three such ranks in the central space, which is but thirty feet wide, and one similar rank in each aisle, we count no less than fifty isolated columns, in addition to the solid piers flanking the central space, and responds in the outer walls, to support the vaulting of an area which is but seventy feet long, by seventy-four feet broad. The clear space between each pair of isolated columns is five feet eight inches from north to south, and five feet three inches from east to west. These very narrow spans, and the general timidity of the design, might well suggest doubts whether this crypt is not of earlier date, in the eleventh century, than 1084. Professor Willis, however, says that the architect of Worcester crypt had certainly seen the crypt at Winchester, and that he has evinced originality and taste of a superior order.†

Such a crypt is like a forest of stone, and its columns form a perplexing labyrinth. When St. Wul

* For these measurements, and for the Table of Dimensions, I am indebted to the courteous kindness of the Rev. Richard Cattley, a minor canon of Worcester, to whom and to Lord Alwyne Compton, the Dean, my thanks are due for helpful information cordially rendered. + Archæological Journal, xx., 91.

stan, in his Life of St. Swithin, describes a crypt built by the Saxon St. Athelwold, at Winchester, I believe that the model from which he drew the picture was such a crypt as this at Worcester. His words respecting crypts are as follows:

66

Insuper occultas studuistis et addere cryptas,

Quas sic Dædaleum struxerat ingenium
Quisquis ut ignotus veniens intraverit illas,
Nesciat unde meat quove pedem referat."

These lines have been frequently quoted, or referred to, as proving that the Saxons had large crypts, but they do nothing of the kind. The narrow spans of the early crypts, necessitating the use of a multiplicity of columns within a very small space, were the actual sources of the labyrinthine perplexities experienced by strangers, as described by St. Wulstan. In the crypt at Worcester, Bishop Wulstan held an ecclesiastical Synod in 1092, and we can imagine that the members present were more puzzled by the difficulty of finding their way, amidst its many closely-placed columns, than by any arguments propounded at the Synod.

to

Serlo, Abbot of Gloucester, began, in 1089, rebuild his church, which had been burnt down in 1087. The Gloucester crypt, which may thus be dated circa A.D. 1089-91, shews a very considerable advance in boldness of design, upon that of Worcester. It is longer, but of about the same width, yet the number of isolated columns in the Gloucester crypt is but twelve (in place of fifty at Worcester), and its aisles are divided from its central space by arcades, each of four arches, on massive piers. This lessening of the labyrinthine character, of the

* Wolstani Liber de Vita S. Swithuni, Acta Benedictin., sæculum

V., p. 630.

crypt's columns, arises from the bolder design of the Abbot of Gloucester, who adopted wider spans for his vaulting. From Mr. F. S. Waller's plan of this crypt,* we learn that between pairs of vaulting columns (of which there are only two ranks, each containing six columns, in the central space), there is a clear space of ten feet from north to south, and of seven feet from east to west. The vaulting spans in the aisles of the crypt are still wider, and evince great boldness of design, when compared with the spans used at Worcester. In plan, and in number of alleys, the Gloucester crypt was very similar to Ernulf's crypt at Canterbury; but in length, height, width, and spans of vaulting, Ernulf's excels it.

Casing, and additional masonry added to the Gloucester crypt at subsequent periods, have given it a patched and uncomely appearance. It has three apsidal chapels.

At Winchester, Bishop Walkelin, as the records state, began to rebuild his church, from the foundation, in A.D. 1079. His crypt, beneath the choir, has two aisles, flanking a central space, thirty-four feet wide, which is divided by a single row of five columns,† It has but four alleys, in a total width of eighty-three feet. The boldness of its plan, and the wide spans of its vaulting, should have prevented any antiquary from supposing that this crypt was built about A.D. 980, by St. Athelwold. It is utterly unlike the crypts described by St. Wulstan. Yet such a mistaken notion was generally received, until very recent times.

*To Dr. Law, the Dean of Gloucester, I am much indebted for his kindness in sending to me a copy of Mr. F. S. Waller's paper upon and plan of the crypt of that Cathedral.

† Archæological Journal, xx., 90. There is a later crypt eastward of this; but it is narrower, and has but two alleys, formed by a rank of four columns.