the Saxon fortress. What caused the change from G to C is hard to
discover, but when King John was on the throne the name was written
Cantebrige, and the "m" put in its appearance in the earlier half of
the fifteenth century, the "t" being discarded at the same period. It
seems that the name of the river was arrived at by the same process.
Perhaps the oddest feature of the whole of these vicissitudes in
nomenclature is the similarity between the Roman Camboritum and
Cambridge, for the two names have, as has been shown, no connection
whatsoever.

A map of Cambridgeshire, compiled by the Rev. F.G. Walker, showing the
Roman and British roads reveals instantly that the university town has
a Roman origin, for it stands at the junction of four roads, or rather
where Akeman Street crossed Via Devana, the great Roman way connecting
Huntingdon and Colchester. Two or three miles to the south, however,
the eye falls on the name of a village called Grantchester, and if we
had no archaeology to help us, we would leap to the conclusion that
here, and not at Cambridge, was the ancient site mentioned by the
earlier chroniclers. And this is precisely what happened. Even recent
writers have fallen into the same old mistake in spite of the
discovery of Roman remains on the site of the real Roman town, and
notwithstanding the fact that the two roads mentioned intersect there.
The trouble arose through the alterations in spelling in the name of
the village of Granteceta, or, as it often appears in early writings,
Gransete, but now that Professor Skeat has given us the results of his
careful tracking of the name back to 1080, when it first appears in
any record, we see plainly that this village has never had a past of
any importance, and that the original name means nothing more than
"settlers by the Granta." There is a Roman camp near this village, and
a few other discoveries of that period have been made there, but such