form of canopy may be followed, but care should be taken to avoid the appearance of insecurity, or of what is as bad, dependence upon metal ties, which form no part of the construction as recognised in the design. It must also be large and high, and its eastern arch be closed by a reredos. And the light should be so arranged that the canopy shall not throw a dark shadow over the altar. This is particularly important by artificial light, which does not diffuse itself over an interior like daylight. It would no doubt be considered "correct" to stick a lot of candles on the top of a canopy, but they would do harm unless the altar below were itself sufficiently lighted up to neutralise their effect.

There are several reasons why wood seems to me to be a most suitable material for altar-canopies. We want stability, combined with lightness of appearance, richness, and importance, but at the same time subordination, or, if I may say so, subservience to the altar. I know nothing so well calculated to afford these results as wood, not in logs tortured into the forms of starved masonry, but with its woodenness recognised and respected, as we find it in the English joiners' work of the fifteenth century. It is at once light and substantial, and it is capable of the highest artistic treatment, both in sculpture and painting, and with it we are enabled to use a design of sufficient importance to attract the attention from the most distant parts of a building, whilst the minute division of parts prevents its clashing with the broad mass of the altar below. Such a canopy should, I think, not be carried on four equal posts at

the angles like the marble one, but should grow, as it were, from the reredos behind, supported at the sides by light screen-work, but entirely open in front. Indeed, I cannot point to a better executed model than the canopy over the throne in the House of Lords, which I have already used as an illustration when speaking of the reredos.

Cast metal is a material which also may be used for altar-canopies, either alone or in combination with marble. If it be used alone, the general form may be such as I have just recommended for wood; if in combination, the marble will probably regulate the design. A good canopy cannot be made of wrought metal; it is an exceedingly "correct" material, owing to the ease with which it can be twisted into those extraordinary shapes which the fashion of the day insists upon believing to be "mediæval," and consequently we often see it doing duty for which it is absurdly unfitted. For a grille in which lightness, strength, and unobtrusiveness are required, nothing can be better; but for a broad, monumental treatment, it is by its nature unsuitable, and the modern attempts to use it in this manner have in consequence always proved failures.

It is unnecessary to discuss the iconography of the altar-canopy, as it would be but to repeat what was said concerning the reredos.

SECTION XVII.

OF THE CREDENCE AND PISCINA.

I INCLUDE these in one section, because, properly speaking, the latter is but an appendage of the former. The credence is a sort of side table, upon which certain vessels and other things required during the celebration of the Holy Eucharist are placed and made ready, so as to be conveniently near to the altar when wanted. The piscina is a sink or drain, through which the water used in several ablutions is poured away.

Of all the surroundings of the altar the credence is that of the most practical importance. It is therefore particularly unfortunate that in ninety-nine out of a hundred modern churches it is either omitted or badly placed. There is no point upon which modern churchbuilders are more obstinately "correct" than in placing the credence on the north side of the sanctuary, a position not only exceedingly inconvenient, but what its advocates will probably think more serious-altogether opposed to every authority and "mediæval precedent." The mistake is so general that I hope I may be pardoned for departing for a short time from my rule of excluding purely archæological considerations from my argument, in order to show that the usage of the old builders was

in this matter also in accordance with good sound com

mon sense.

In seeking for information as to what was the ancient practice, we naturally look first to the old churches which still remain to us; and there on the south side of every altar, almost without exception, we find a single or double niche, generally containing a drain (sometimes two), and often a shelf or bracket. Sometimes this niche is so small as to be entirely taken up by the drain, but often it is amply large enough to hold all that is required to be placed on the credence. Not unfrequently, but by no means always, a recess is also seen on the north side of altars. This rarely takes the niche form, and will I think, if examined, always be found to have been provided with doors. In some cases the architectural treatment and decoration prove beyond doubt that it was intended for the reception of the blessed sacrament, but often there is nothing to indicate this, and sometimes it appears to be larger than would be provided for that purpose; and again its occasional occurrence at minor altars argues some less important use. Other recesses and ledges, which might be used as credences, are occasionally found near altars, but the two-the niche with the drain on the south, and the locker on the north-are of such constant occurrence, that if the credence was a fixture forming part of the fabric of the church it must have been one of them. Each has been claimed for it.

The old English service books, so far as I have observed-but I must confess that my knowledge of them

is only slight-give no direct information about the position of the credence. The word credentia does not occur; indeed, this name is of comparatively modern. introduction, and of secular origin. But we have abundant evidence of the existence of the thing, and how it was used in St Osmund's Consuetudinary, chap. xciii. "Post introitum vero misse, unus ceroferariorum panem et vinum et aquam in pixide et phiolis solempniter ad locum, ubi panis, vinum et aqua ad eucharistie ministrationem deponuntur, deferat. Reliquus vero ceropherarius pelves cum aqua et manutergio." And again, “Interim etiam veniant duo ceropherarii obviam acolito ad ostium presbyterii cum veneratione ipsum calicem ad locum predicte administrationis deferente offertorio et corporalibus ipsi calici superpositis. Calice itaque in

loco debito reposito, corporalia ipse acolitus super altare solempniter deponat. Lecta Epistola, subdiaconus panem et vinum, post manuum ablutionem, ad eucharistie ministrationem in loco ipsius administrationis preparet ministerio acoliti."

We learn from these passages that the objects which were placed in the locus administrationis—evidently what we now call the credence-were not kept there permanently, but brought in during the time of the service, which leads to the conclusion that the locker on the north side of the presbytery is not the place referred to, for there would be no use for the doors with which we always find it to have been provided. But we can here find no certain information as to either the form or the position of the old credence. Fortunately, however,