one half of the donativum or imperial largesse,[1] which was employed to provide for those who through sickness or other misfortune had to abandon the service prematurely, while the legionary soldiers at the same time contributed to another fund, which was also kept there and from which was defrayed the cost of their own burial.[2] The standard-bearers thus became the custodians of the regimental funds, and naturally the men who filled the office were chosen not only for their fidelity but also for their education.[3] To protect the sacred emblems and the treasure of the Sacellum a special guard was posted. An inscription from Aquincum commemorates the restoration of a guard-room for the men guarding the standards and the sacred images,[4] while at Ostia the name of M. Mikenius Julius, trumpeter of the Seventh Cohort 'attached to the guard' was found scratched upon a stone near the shrine.[5]

While the Sacellum with its standards and its images has everywhere disappeared, the treasure-chamber in many cases remains. A typical vault was recently discovered within the Principia at Brough in Lancashire. It was eight feet in length, and varied in width from five to seven feet, with a depth of eight feet. It was entered at one end by a flight of eight steps. The floor was of cement. In the wall, broken up and used as a building stone, was a portion of an inscribed slab dating from about the year A.D. 158, showing that the vault itself was not earlier than the second half of the second century.[6] At Lambaesis the treasure chamber was about six feet in depth. At Aesica, Chesters, Bremenium and South Shields it was also of some depth. On the other hand, at Wiesbaden,[7] at Gnotzheim,[8] and at Murrhardt,[9] its foundations lay near the surface, indicating that at these forts, as at Newstead, the floor of the Sacellum itself must have been raised considerably above the natural surface of the ground, and must have been approached by a flight of steps. It seems probable that in many instances such treasure-chambers were added in the latter part of the second century. Thus the vault at Brough, as we saw above, evidently belongs to a reconstruction of the Principia. This seems to be the case also at Butzbach,

1 Vegetius, 2. 20.

2 Cagnat, L'Armée Romaine d'Afrique, p. 457; also Nicole, Archives militaires du 1er Siècle (Geneva, 1900), p. 18.

3 Vegetius, 2.20.

4 Von Domaszewski, op. cit. p. 13.

5 Ibid. p. 14.

6 Haverfield, Victoria County History: Romano-British Derbyshire, p. 204.

7 Der Obergermanisch-Raetische Limes, Lief. 31, Kastell Wiesbaden, p. 25.

8 Ibid. Lief. 70, Kastell Gnotzheim, p.8, Taf. ii. fig. 2.

9 Ibid. Lief. 44, Kastell Murrhardt, p. 7.