184 THE ESSEX NATURALIST
Our experience, therefore, in Essex and in Hertfordshire, has consistently
shown this trail of related boulders to be unique. Nowhere have we dis-
covered branching trails or duplicated tracks, and in every instance the
boulder of the track has an undoubted antiquity. The random distribution
of Puddingstones in Essex and Herts has emphasised, rather than confused,
the reality of the track.
2. THE EVIDENCE OF THE PAGAN STONES
No less than 1G of the boulders of the alignment are incorporated in the
fabric, or lie in the vicinity of, a pre-Conquest Church. Eight of these are
in Essex alone, viz.: Epping Upland, Magdalen Laver, Beauchamp Roding,
Broomfield, Great Leighs, Fairstead, Marks Tey and Fordham. In
addition to these, the line passes through three other pre-Conquest sites—
at Berners Roding, Chignal and Rivenhall, where, up to the present, no
stone has been located.
The reasons for the inclusion of "pagan" stones in early Christian
churches are sufficiently well-known, and we may confine our remarks to
the general picture in Essex in so far as it affects our case. According
to the Inventory of the Royal Commission on Historical Monuments, there
are close on 60 pre-Conquest churches in Essex. Other writers put the
total somewhat higher than this, and therefore, to be on the safe side, we
have examined upwards of 190 of the 390 parish churches in the county.
The results were surprising. Not more than 20 churches were associated
with a boulder of any kind which could in the widest sense be regarded as
a "pagan" stone. Of these, nine were Puddingstones, and the rest sarsens;
and, of the nine, eight were at the pre-Conquest churches upon the align-
ment listed above. The ninth, the only exception, was at East Horndon,
where a puddingstone lies with a heap of debris, including a sarsen, beside
the porch. This distribution of "pagan" stones is given in Figure 3, which
indicates that the incidence of sarsens is not confined to any one district.
On the other hand, it strikingly emphasises the coincidence of the con-
glomerate "pagan" stone group with the general alignment.*
The "pagan" stone must not he confused with stones and boulders
employed in general construction. In a county singularly devoid of sources
of building stone, it was the practice in early times to make use of local
material of many kinds—flint rubble, Septaria, clunch, sarsens, etc. In
some areas a soft ferruginous conglomerate abounds in the river valleys,
and we have listed at least 35 churches in which this material has been
utilised to a greater or lesser extent. In contrast to this, the one "pagan"
boulder, often incongruously protruding from the lower courses of the wall,
cannot be mistaken (Plate 7). The phenomenon is found throughout the
alignment, from Norfolk to Buckinghamshire, e.g., at Heacham, Ingoldis-
thorpe and Snettisham, in Norfolk, and, most remarkable of all, at
Chesham, Bucks, where the church appears to rest upon a puddingstone
circle. We know of only two puddingstone inclusions at all comparable
with these of our series—at St. Stephen's, St. Albans, and at Caddington,
near Luton.
* A tenth "pagan" puddingstone has since been discovered in the
foundation wall of the pre-Conquest church at Stifford.