D B Campbell The Roman Army in Detail: The Problem of the First Cohort

[Renatus]

However, I am getting a little concerned about the number of men on guard duty as it starts to seem a little excessive

Vegetius tells us that the night was divided into four watches, so that none of the guard had to be awake for longer than three hours. Thus, only a quarter of the guard was on watch at any one time. This alters the spacing of those manning the ramparts. The remainder would have been resting, armed ready to repel any surprise attack. This raises the question of what shelter, if any, was provided for them. As Hyginus' treatise is essentially theoretical, his thinking may not have extended to that level of practicality. However, there is another possibility. The via sagularis in Hyginus seems to run parallel to the intervallum but those who have studied the layout of permanent forts and fortresses tend to equate the two. De Voto translates via sagularis as 'Cover Avenue' but, if one has to translate it (and personally I prefer not to translate technical terms), 'Cloak Street' might be more accurate. One could, therefore, imagine those of the guard who were not actually on watch resting on the via sagularis behind the rampart, wrapped in their cloaks. If this is not too fanciful, it might explain the rather peculiar naming of that part of the camp.

Why I am a bit partial to 8 men per century on guard duty is the Romans do have a passion for the principal of one tenth.

The problem with this is that it requires the substitution of nine tents for the eight that, in this instance, is quite clear in the text. This smacks of altering the source to fit the theory. More likely is that, to make up his numbers for the guard, Hyginus simply doubled up on the normal principle.