The chief roads (mainly British trackways) are shown with broken lines. Each infantry block roughly indicates 5,000 men. The whole country north of the Thames was hostile to the Romans, perhaps much of that to the south also. The Roman troops at Lindum (Lincoln) were only the defeated remnant of the IXth Legion. It should be noted that of the settled and corn-growing districts, one was occupied by the Britons, and there remained open for Paulinus’s army, and the crowd of refugees which accompanied it, the choice between Kent or the Lower Severn Valley. The IInd Legion at Isca Silurum (Caerleon) had orders to march on Londinium, and Paulinus would expect it to be well on its way. The heavy black line stretching south-west from London indicates the probable direction of the Roman retreat.] The object of Paulinus was to complete his frustrated combination. At Deva, two hundred miles away, were perhaps 5,000 men; at Lindum, a hundred and thirty miles to the north-east, perhaps an equal force, dispirited by defeat. If he took the road to Deva, or that to Lindum, he would have the Britons upon him. Is it conceivable that this able general, with supply difficulties aggravated by his mass of non-combatants, would deliberately plunge into the midst of the enemy, in order to join one of his two smaller detachments, when in the Lower Severn Valley lay a whole legion and its auxiliaries. If his orders were being obeyed, it should be already on the march; but if it had not yet concentrated, its nearest detachments were only a hundred miles away. A study of the map will show that, if London were abandoned, Corinium (Cirencester) would be the natural point of concentration for Paulinus’s army, the IInd Legion, and the garrisons of Deva and Viroconium. The troops round Lindum and the garrisons in Kent must, for the moment, be left to themselves. We are justified in thinking that Paulinus would move in the direction of his largest outlying corps--the IInd Legion. Considerations of supply would also take him westward. Food might be found in Kent, but not reinforcements. The conclusion is that, for every reason, the direction of the retreat would be westward. Paulinus no doubt crossed the Thames, presumably by the bridge at London, which would, of course, afterwards be destroyed, and retreated towards Calleva (Silchester). It is probable that Verulam was taken and sacked by the Britons after Paulinus had passed through it. Tacitus only says that it fell at about the same time as London. The British host then moved on to London, which shared the fate of Camulodunum and Verulam. The massacre here was probably the worst, for it would, naturally, apart from its commercial importance, be full of fugitives.