Page:VCH Warwickshire 1.djvu/293

 ROMANO-BRITISH WARWICKSHIRE evidence is not convincing. Neither discoveries of remains, nor the local nomenclature, nor the physical features of the country really aid us. No Roman remains have been found in Birmingham except a few coins (p. 244), and coins help little in such a case ; so far as they go, however, they favour a line east of Hutton's and nearer the direct line mentioned at the outset of this paragraph. A piece of ancient road was discovered about 1870 or 1875, near Chad Valley House in Westbourne Road, Edgbaston, and Mr. J. A. Cossins, who saw it, has told me that it was 5 feet underground, paved with large pebbles of local gravel, and was not in line with the commonly supposed direction of the Roman road. A well near Metchley, a bit of old road near Harborne Park Road, and some horseshoe draining tiles found in January, 1902, have all been called Roman, without the slightest reason. Nor do local place-names help us. Icknield Port Road is unquestionably a modern invention, and the title Icknield Street, as applied to the road connecting Hunter's Lane and Monument Lane, is not demonstrably old. Negative evidence is, of course, imperfect ; but I cannot trace the title back beyond 1825, and in 1553 a part, at least, of this road seems to have been called the Slade. The title Icknield Street may therefore have been introduced as a result of Hutton's theory. Certainly, if old names are to be quoted, Holloway Head should not be forgotten, though that would favour rather the direct line indicated in the third sentence of this paragraph. Nor again is it possible, amid the vast developments of a great city, to reconstruct the original hills and valleys and judge whether they were such as to divert a Roman road from its straight course. That kind of judging is always a dangerous speculation ; in this case it is best omitted wholly. After all, the straight course outlined at the commencement of this discussion is the simplest, and in default of other reasons the least improbable. Here we must leave the problem unsolved. It is not inappropriate that a characteristically modern city should have lost for ever the recollection of her most ancient road. There remains another problem, almost as difficult as that which we have just dismissed. For convenience we have called the road Rycknield Street : we have now to trace out thie tangled history of that name. We start from the similar name Icknield. Icknield Street, properly so called, is an ancient trackway through Berkshire and Oxfordshire, of which the course is still visible, and the name, under the form of Icenhylt or Icenhilde Street, is attested in documents earlier than the Conquest. It is not a Roman, but perhaps a British road, and so far we have here no concern with it. But we are concerned with its name. For when the antiquaries of the twelfth and following centuries began to treat of the so-called ' Four Roads,' they got hold of the name Ick- nield, obviously without knowing what exactly it meant. One of them said that it ran from east to west which is roughly true and another said that it ran from north to south. This latter was identified with our road ; not, so far as we can tell, because of any local name, certainly not because of any Iceni in the west, but probably because this road alone i 241 31