Towards the end of the fourth century, the naval expeditions of the Saxons attracted the notice and excited the fears of the Britons and the Gauls: their vessels apparently were unfit for a long voyage, or for encountering either the dangers of the sea or of battle; they were flat-bottomed and slightly constructed of timber, wicker-work, and hides; but such vessels possessed advantages, which to the Saxons more than compensated for their defects: they drew so little water that they could proceed 100 miles up the great rivers; and they could easily and conveniently be carried on waggons from one river to another. We have already noticed the itineraries of the Roman empire: of these there were two kinds, the _annotota_ and the _picta_; the first containing merely the names of places; the other, besides the names, the extent of the different provinces, the number of their inhabitants, the names of the mountains, rivers, seas, &c.; of the first kind, the itinerary of Antoninus is the most celebrated: to it we have already alluded: to the second kind belong the Peutingarian tables, which are supposed to have been drawn up in the reign of Theodosius, about the beginning of the fifth century, though according to other conjectures, they were constructed at different periods. The beginning of the tables is lost, comprising Portugal, Spain, and the west part of Africa; only the south-east coast of England is inserted. Towards the east, the Seres, the mouth of the Ganges, and the island of Ceylon appear, and routes are traced through the heart of India. Dr. Vincent remarks, that it is a very singular circumstance that these tables should have the same names in the coast of India as the Periplus, but reversed. Mention is also made in them of a temple of Augustus or the Roman emperor: these circumstances, Dr. Vincent justly observes, tend to prove the continuance of the commerce by sea with India, from the time of Claudius to Theodosius; a period of above 300 years. In these tables very few of the countries are set down according to their real position, their respective limits, or their actual size.