curtailing the extent of the area occupied by the cross, may have been designed to lessen the liability of the latter to collisions with market carts. It is supposed that the top of the central spire fell in the eighteenth century, damaging the substructure. Anyhow, at some time in the seventeenth, or in the early part of the eighteenth century, the hexagonal shelter was taken down from around the central pier (which still remains intact), and was then rebuilt in its present form, portions only of the old Gothic parapet, and the pinnacles, being re-used. This rebuilding has escaped record, but that it did take place the internal evidence of the structure itself makes sufficiently obvious. The absence, already mentioned, of buttresses; the clumsy, square blocks which do duty for the bases of the piers; the classic imposts of the latter, and the depressed arches (unconstructional, because they are not turned with voussoirs, but formed each of one huge pair of stones, cambered to simulate an arch in outline), and the exaggeratedly prominent keystones, could never have been perpetrated at the early date of 1500, but at some subsequent rebuilding, of which the sum of them affords cumulative and convincing proof. Charles Pooley (_Old Stone Crosses of Somerset_, 1877) states that the cross was rebuilt from the ground in 1841: but he was clearly mistaken. Dr F. J. Allen, of Cambridge, is positive on this point. His grandfather, as one of the trustees of the Shepton Mallet cross, was largely responsible for the rebuilding in question; and his own mother and uncle, living as children in their father's house, facing the cross, were eye-witnesses of the progress of the work, and could distinctly remember that only the spire above the roof was reconstructed. Minor repairs may have been done at the same time to the rest of the building, but it was certainly not taken down bodily. The architect employed was G. B. Manners, of Bath; and it is claimed that his design for the modern spire is a careful reproduction of the original one. To what extent this is the case may perhaps be judged by comparing the spire actually standing with an illustration, which appeared in the _Gentleman's Magazine_, 1781, from a drawing made in 1747. The latter may be faulty, but, such as it is, its value as a record can scarcely be overrated, since it furnishes the earliest extant version of Shepton Mallet cross. The accompanying letterpress says: "On the top of the cross, on the east side, are figures in niches, and, above all, a modern weathercock." The engraving, it is true, shows figures on more sides of the head than one; but the discrepancy need not be material, if one may conjecture that all the figures, other than those on the east side, had perished in the interval between 1747 and 1781. In any event the massive, carved stone cylinder, depicted as capping the spire in 1747, cannot have been the original cross-head of 1500, which, according to Pooley, was "a heavy,