70
THE ESSEX NATURALIST
where, for example, only one alignment is predicted, and
where I find either none or two. The agreement is usually
within 10 per cent.
With this rather long preliminary explanation, I will now
give in Table 3 the results which I have obtained by plotting
all the ancient parish church sites, the castles and tumuli
in each of three areas of Essex.
Table 3
Alignments of Ancient Sites in Essex.
Order of Calculated for Alignments
Alignment Random Distribution found
S. W. Essex. 3-spot ... (46.6) 47 ... 47
4-spot ... (3.9) 4 ... 5
5-spot ... (0.3) 1 ... 1
N.E. Essex. 3-spot ... (71.(3) 72 ... 78
4-spot ... (7.8) 8 ... 11
5-spot ... (0.8) 1 ... 2.
Mid-S. Essex. 3-spot ... (89.1)89 ... 113
4-spot ... (9.7) 10 ... 8
5-spot ... (0.95) 1 ... 2
6-spot ... (0.09) 0 ... 1
7-spot ... (0.0081) 0 ... 0
It will be seen that in the first two areas the alignments
found are in fair agreement with the calculated values for
random distribution. I have re-checked my preliminary
determinations (made on tracings from folded 1" O.S.
sheets) by using flat, uncoloured sheets bought for the pur-
pose and carefully preserved from paper distortion, and Table
3 shows the most recent results.
I must here interpolate a warning to those who may be
tempted to seek very long alignments of sites plotted on a
map; you may know that map-makers resort to several
devices in order to represent on a flat sheet of paper the
features of the Earth's curved, spheroidal surface. Whatever
kind of map-projection may be used, there is inevitably some
distortion of direction or of size in making a map of the
country. In the case of the recent Ordnance Survey maps