Stow says the Thames rises in Winchcombe in Gloucestershire, but this is not really the case.
Harrison in his " Description of Britain " describes its source as in the side of an hill in the plains of Cotswold, about one mile from Tetbury, but the stream is so small, being often dried up in the summer, that the assertion is disputed and its source is said to be from the pool above Kemble.
Camden gives the source as a copious spring called "Thames head" near the village of Tariton, about two miles south-west of Cirencester. This corresponds with the modern
source of the Thames" as shown in the Atlas of the English Counties. Camden says another source suggested is at Cobberley (Cubberly), at a place called Seven Wells Head, but that the former opinion is the more prevalent (Britannia, I. 412, ed. Gough).
It appears from the map that Seven Wells Head is the source of the Churn, a tributary of the Thames. It is interesting to note that there is a place called Witcomb about two miles from Seven Wells Head. May it be that Stow was confusing Winchcombe with Witcomb and inclined to the " Seven Wells head " source rather than to the " Thames head" source, the theory advanced by Camden?
The river flows through London to its estuary, and London owes its importance as a centre of trade and commerce to its situation upon the banks of the river.
As to the origin of the name, the "h" in the word is unoriginal. Cæsar has "Tamesis" Tacitus, " Tamesa" ; O.E. forms, " Temese," " Temes," " a " being mutated to "ae" and "e" ; M.E., "Temese," "Temse," "Tempse," and in some MSS. "Themese." The English habit of dropping the vowel of the second syllable in trisyllabic names reduced "Temese" to "Temse" and led to insertion of "p." Final "e" dropped off, leaving" Tems." Misspelt in later times with" a " because so spelt in Cæsar and Tacitus, but the pronunciation preserves the true form of the name. "H" due to A.N. scribes (Skeat).
Roman and other coins were found in the bed of the river during excavations for the new London Bridge, 1824 (Arch. XXV. 600), and bronzes, 1837 (R. Smith, 67-8, 72).
The marshland from the river's edge extended about 300 feet inland, shelving up towards Thames Street (ib. 601).
It seems probable that the level of the Thames in old days was 12 feet lower than at present.