I have suggested how these might have developed from the tump, and shown where pond and tump were used together. Moats are a similar arrangement on a larger scale. The trackways go straight for the island part of the moat. It is not the least amazing part of
this revelation that I find practically all the small horse or cattle ponds in field or homestead which are marked on a 6in. ordnance map have leys running through them, and that examination in dry seasons shows signs of the road passing through them. "And when we cleaned the pond out we found it cobbled at the bottom" is a frequent report made by a farmer. I show a photograph of one of these at Bridge Sollars, with the trench of the road beyond.
A beautifully constructed causeway of even pitched stones is to be seen at the foot of Holmer Hill (Plate VI.). It has well defined edges, and lies at the bottom of a small sighting pond. In the crevices of its stones I found fragments of crude red pottery, with a bit of early (Anglo-Saxon) ornament, a bit of iron slag, and a bit of iron. This ley is sighted on the North Hill, Malvern.
I cannot say that passengers walked through the bottom of these ponds (most of them have one shelving edge, with the opposite bank steep), but to this day an ancient road (at Harley Court, Hereford) does go through the bottom of a small pond, being sighted through the Cathedral.
When there is a large central island on a moat I surmise early dwelling houses--a subject for spade research. There evidently came a wish for roads not running through the water, and a pair of ponds or lakes with a causeway between, such as we find at Holmer fish ponds, is frequently found on the map, and is the sure indication of an ancient trackway. Probably the square moats are later than the circular ones. I saw in the grass the track of a 15-foot road (probably Roman construction) making straight for the centre of Yarkhill Moat.
Many ponds (as at Belmont, The Burcot, and adjoining Ledbury Churchyard) not known as moats are really such, their islands being sighting points.
The causeway to the centre of the moat evidently suggested their use (many ages after they were made) as a defensive ring of the house of a rich owner, as at Brinsop, Badesley Clinton, Gillow, etc.
I think that the word lake, now used for large sheets of water, was originally applied to small reflecting sighting ponds as well. The place names of Sutton Lakes, Withington Lakes, Letton Lakes, and Tumpy Lakes are explained by this theory.