I will quote the evidence, such as it is, and the reader will perceive how difficult it is to arrive at a conclusion.

Of actual Roman remains all we have is a couple of coins of the end of the fourth century (probably minted at Constantinople), a silver ingot of the same period, and a funeral inscription.  No indubitably Roman work has been discovered.

On the other hand there has been no modern investigation of those foundations of the White Tower where, if anywhere, Roman work might be expected.  This exhausts the direct evidence.  In sciences such as geology or the criticism of Sacred Books evidence to this extent would be ample to overset the firmest traditions or the most self-evident conclusion of common human experience.  But history is bound to a greater caution, and it must be reluctantly admitted that the two coins, the ingot and the bit of stone are insufficient to prove the existence of a Roman fortress.

Leaving such material and direct evidence we have the tradition, which is a fairly strong one, of Roman fortification here, and we have the analogy, so frequently occurring in space and time throughout the history and the area of Western Europe, that Gaul reproduces Rome.  What the Conqueror saw (it might be vaguely argued) to be the strategical position for London, that a Roman emperor would have seen.  But against this argument from tradition, which is fairly strong, and that argument from analogy, which is weak, we have other and contrary considerations.

Rome even in her decline did not build her citadels outside the walls:  that was a habit which grew up in the Dark and early Middle Ages, and was attached to the differentiation between the civic and military aspects of the State.

Again, Roman fortification of every kind is connected with earthworks.  So far as we can tell from recorded history the ditch round the Tower was not dug till the end of the twelfth century.  Finally, there is this strong argument against the theory of a Roman origin to the Tower that had such a Roman fortress existed an extension of the town would almost certainly have gathered round it.

One of the features of the break-up of Roman society was the enormous expansion of the towns.  We have evidence of it on every side and nowhere more than in Northern Africa.  This expansion took place everywhere, but especially and invariably in the presence of a garrison, and indeed the military conditions of the fourth century, with its cosmopolitan and partially hereditary army, fixed in permanent garrisons and forming as it were a local caste, presupposed a large dependent civilian population at the very gates of the camp or stronghold.  Thus you have the Palatine suburb to the south of Lutetia right up against the camp, and Verecunda just outside Lamboesis.  Now there is nothing of the sort in the neighbourhood of the Tower.  It seems certain that from the earliest times London ended here cleanly at the wall, and that except along the Great Eastern Road the neighbourhood of the Tower was agricultural land.