0.5

1 cm

THE ANTONINE WALL REPORT.

129

If we knew how many horizontal rows of sod there were in a

section of the vallum in its perfect state, it is obvious that that

number multiplied by the original thickness of the sod would

give the approximate height. But it is a hard question to

answer how many layers of sod there were in the vallum. The

largest series yet found was at Roughcastle No. 3 section,

where there were 19 layers now standing, 4 feet 8 inches high,

or an average of 3 inches to each layer. If we could assume

that the original sod wTas not less than 5 inches thick,1 and that a

few layers have by time and weathering disappeared, we might

with tolerable safety infer that (independent of any wooden battle-

ment) the wall at Roughcastle was not much over 10 feet high.

That section at Roughcastle, however, is not a portion of the

wall proper (as it is the wall of the camp), but at Croy No. 11 and

Barr Hill No. 1 we have sections of the actual vallum which in

height resemble that at Roughcastle. In these there are 19 layers

with heights of 58 and 52 inches. Making similar allowances to

those postulated at Roughcastle, the heights of the Croy and Barr

Hill portions might be reckoned also as not exceeding 10 feet. The

evidence of the sod layers, therefore, so far as it goes, would not

warrant a belief that the wall was ever much higher than 10 feet.

One point of great importance in this problem appears to have

escaped the notice of all previous writers. The sides of a wall

of earth, whether of sods or not, must have had a decided

" batter "—that is, in order to preserve stability, it must have been

necessary to narrow the breadth as the structure rose, so that

the top of the wall would be considerably narrower than the

base. It is very interesting to find in the work of the Roman

architect Vitruvius 2 a canon for the breadth of a rampart. It is

true that he refers to the wall of a fortified town, but the dis-

tinction between such a wall and a frontier rampart, whether

erected primarily for defensive purposes or not, can scarcely

have been vital. " The thickness of the wall," he says, " should be

1 Half a foot was, it will be remembered, the theoretical thickness for the

Roman sod (p. 30, supra), but that standard could not be attained in average

soils in Scotland.

-Vitruvius, i., cap. 5. " Crassitudinem antem muri ita faciendam censeo, uti

armati homines, supra obviam venientes, alius alium sine impeditione praeterire

possint."

I

129

If we knew how many horizontal rows of sod there were in a

section of the vallum in its perfect state, it is obvious that that

number multiplied by the original thickness of the sod would

give the approximate height. But it is a hard question to

answer how many layers of sod there were in the vallum. The

largest series yet found was at Roughcastle No. 3 section,

where there were 19 layers now standing, 4 feet 8 inches high,

or an average of 3 inches to each layer. If we could assume

that the original sod wTas not less than 5 inches thick,1 and that a

few layers have by time and weathering disappeared, we might

with tolerable safety infer that (independent of any wooden battle-

ment) the wall at Roughcastle was not much over 10 feet high.

That section at Roughcastle, however, is not a portion of the

wall proper (as it is the wall of the camp), but at Croy No. 11 and

Barr Hill No. 1 we have sections of the actual vallum which in

height resemble that at Roughcastle. In these there are 19 layers

with heights of 58 and 52 inches. Making similar allowances to

those postulated at Roughcastle, the heights of the Croy and Barr

Hill portions might be reckoned also as not exceeding 10 feet. The

evidence of the sod layers, therefore, so far as it goes, would not

warrant a belief that the wall was ever much higher than 10 feet.

One point of great importance in this problem appears to have

escaped the notice of all previous writers. The sides of a wall

of earth, whether of sods or not, must have had a decided

" batter "—that is, in order to preserve stability, it must have been

necessary to narrow the breadth as the structure rose, so that

the top of the wall would be considerably narrower than the

base. It is very interesting to find in the work of the Roman

architect Vitruvius 2 a canon for the breadth of a rampart. It is

true that he refers to the wall of a fortified town, but the dis-

tinction between such a wall and a frontier rampart, whether

erected primarily for defensive purposes or not, can scarcely

have been vital. " The thickness of the wall," he says, " should be

1 Half a foot was, it will be remembered, the theoretical thickness for the

Roman sod (p. 30, supra), but that standard could not be attained in average

soils in Scotland.

-Vitruvius, i., cap. 5. " Crassitudinem antem muri ita faciendam censeo, uti

armati homines, supra obviam venientes, alius alium sine impeditione praeterire

possint."

I