The _pallium_, when used by ecclesiastical writers in its proper and restricted sense, denotes an ornament specially appropriated to archbishops. Its earliest form is shown in the Ravenna mosaics—that of a narrow strip of cloth, passed over the left shoulder, looped loosely round the neck, and then passed over the left shoulder again, so that the two ends hang free, one in front, the other behind. This method of disposition seems to indicate an identity of origin with the _orarium_; indeed, it is sometimes difficult to distinguish between these vestments in early representations. A desire for symmetry, probably, decided the next step in its evolution; this consisted in bringing the free end to the middle and knotting it into the lowest point of the loop: this we find exemplified in monuments of the eighth, ninth, or tenth century. From this the transition to the form which became universal in later times was easy, and the two are found contemporaneously. The final form—which will be more fully described in the third chapter—is that of an oval loop with a long tail pendent from its ends, so that when the ornament is in position it presents the appearance of a capital Y on the front and on the back. The early history of this vestment is involved in deep obscurity. As already hinted, it is not improbably a modification of the _orarium_; but there is no evidence, further than general outward resemblance, that this is actually the case; nor is there any apparent reason for its appropriation to archbishops. The question must remain open till further research either reveals the missing links in the chain of connection, or elicits some more satisfactory solution of the question. The idea of Dr Rock, according to which the _pallium_ is viewed as 'the true and only representation of the Roman toga,' is most unsatisfactory. He thinks that the toga, which was folded over the left shoulder, under the right arm, over the right shoulder, and again over the left shoulder, 'dwindled down to a mere broad band,' folded much the same way; and that this broad band was the early _pallium_. The evolution here supposed is, however, most unnatural; there is not time for it to have taken place between the institution of Christianity and the date of the Ravenna mosaics—much less between the time when ecclesiastical vestments and their development began to receive special attention and the latter date; the toga, as we have already seen, was itself practically obsolete when Christianity began to make itself felt, and still further removed from the current fashion of the time at which archbishops began to require distinguishing insignia; and, lastly, the connecting links between the blanket at one end and the narrow strip of cloth at the other, which Dr Rock adduces and figures, are too few in number to be convincing, and quite explicable on other grounds, such as the unskilfulness of the ancient artist—a fruitful source of error in archæological research.