Fashion in dress or ornament is subject to constant changes which, though perhaps individually trifling, in time amount to complete revolutions; but the devotees of any religion, true or false, are by nature conservative of its doctrines or observances. Combined with the conclusions at which we have just arrived, these two universally recognised statements yield us presumptive evidence of the truth of the theory which views the Roman civil dress as the true progenitor of mediaeval ecclesiastical costume. We have seen that at first the worshippers wore the same costume both at worship and at home. Fashion would slowly change unchecked from year to year, while ecclesiastical conservatism would retard such changes so far as they concerned the dress worn at Divine service: small differences would spring into existence between everyday dress and the dress of the worshipper, and these differences, at first hardly perceptible, would increase as the process went on, till the two styles of costume became sharply distinguished from one another. Parallel cases are not wanting to show that this is not altogether mere random theorizing. For example, the ministers of the Reformed Church of Holland maintained, till comparatively recently, a picturesque fashion of dress over a century old, which they wore only when conducting Divine service.[13] Perhaps, however, the objection may be urged against this view of the case, that if the process were such as we have described, it should apply as well to the worshippers as to the minister: that they, as well as he, should wear service-robes. It is possible that this would actually have been the case had the church services maintained their most primitive form, as St Paul describes it in the First Epistle to the Corinthians: 'When ye come together, every one of you hath a psalm, hath a doctrine, hath a tongue, hath a revelation, hath an interpretation';[14] that is, had all the worshippers maintained an equally prominent position instead of selecting one of their number to conduct their services. As it was, the outstanding position of the minister rendered his equipment especially liable to such stereotyping as we have imagined. In the following chapter we shall submit the truth of this theory to a test. If the genesis of ecclesiastical vestments actually took place in some such manner as this, then the vestments as we find them described in the earliest writers ought to bear conspicuous points of resemblance to the civil costume of the Roman people during the first three Christian centuries. We shall now inquire whether this be so. [1] Mishneh Torah, VIII, section _de vasis sanctuar._, viii 19, where some other particulars are to be found regarding the textures of which the Jewish vestments were made, etc.