was carried on mules and horses. Rüstow supposes that the sutlers and traders only had waggons, with which they followed the army. But the artillery and heavy engines must have been also conveyed in waggons, if they were conveyed at all from one place to another. The want of roads in some countries would probably prevent the use of any carts or waggons, either by the legion or the camp-followers, and every thing that the soldiers could not carry must have been put upon beasts. The soldiers' tents were a chief part of the incumbrances of a Roman army. They were made of prepared skins, in fact of leather. The supply of skins and hides for the Roman army was one of the requisitions laid on the provincials. Sicily, says Cicero, among other things, supplied Rome with skins; and we read, at a later time, of Drusus imposing on the Frisii a fixed contribution of cowhides for military use. Besides the tents, there would be tent poles to carry, with food, utensils, tools, and a variety of things. By making certain assumptions as to the capacity of horses and mules for carrying, and also assuming the weight of the tents and other incumbrances, it is possible to approximate to the number of beasts required for a legion, including the camp servants (calones), and the riding-horses and packhorses of the tribunes. Rüstow thus arrives at a sum total of 520 beasts for each legion. When an army went on a short expedition no beasts would be used, and every soldier carried all that he required.

As the cohort was the unit of the legion, even in the time of Marius, it is probable that his way of drawing up the troops for battle did not differ from that used by Caesar, which we know. The cohort was made by the combination of three manipuli. The legion was formed for battle by placing the cohorts in two, and sometimes in three lines behind one another, with intervals between the cohorts, when the ground allowed it. When there were three lines (acies triplex), three cohorts were placed in front; the three in the second line were placed respectively opposite to the intervals in the first line; and the three cohorts in the third line, as we may conjecture, were opposite to the intervals in the second line, or, in other words, opposite to the cohorts of the first line.

So far as I have described it, this may be a tolerably correct statement of the constitution of a Roman army in the time of Marius. This description is founded on the authorities quoted at the head of this chapter, and a comparison of them with the passages of the ancient writers. As we have no historical writings for the period of which we are now treating, there is nothing to say about the movements of the Roman armies in the field beyond the vague and general expressions of the extant authorities. It is only when we come to Caesar's proconsulship of Gallia, and the wars which followed his invasion of Italy, that we have to deal with the writings of a man who tells us what he did and how he did it.

CHAPTER III.

THE WORKS OF MARIUS ON THE RHONE.

B.C. 104-103.

THOUGH We do not know where Marius was, or what he was doing in B.C. 104 and в.c. 103, we may conjecture that he was in the Provincia, or south of France, part of this time at least, and well employed. Plutarch supposes that he entered Gallia by one of the passes of the Alps, but this statement cannot be relied on. It is more probable that he carried his troops by sea to the French coast. The supplies for his army were certainly taken that way, and brought to the Rhone. The mouths of this river at that time were choked with mud and sand, and the entrance was thus made difficult, and too shallow for the Roman vessels. Marius employed his men in cutting a new channel from some point on one of the outlets of the Rhone to a convenient place on the coast which had water enough for large vessels, and was safe against the wind and waves (Plutarch).

The ancient geographers did not agree about the number of the outlets of the Rhone, and Polybius found fault with Timaeus for saying that they were five. Polybius reckoned only two, and Artemidorus three. Strabo agrees with Plutarch in assigning to Marius the credit of improving the navigation of the river; for, observing that the outlets were choked with the alluvium brought down by the stream, and that the entrance was difficult, he made a new cut, into which the greater part of the water was diverted. There was beyond (east of) the outlets of the Rhone a salt lake, called Stomalimne, which contained abundance of shell-fish and other fish. Some geographers, says Strabo, reckoned the outlet of this lake as one of the mouths of the Rhone, and

especially those who made seven outlets of the river, or five outlets, as the passage of Strabo stands in the present corrected text. But Strabo does not admit this outlet of the lake to belong to the river, for, he says, there is a hill which separates the lake from the river. The Stomalimne is supposed to be the modern Étang de l'Estouma, which lies between the east outlet of the Rhone and the great Étang de Berre, supposed to be the Mastramela Stagnum. "There does exist, in fact, a chain of hills to the west of the gulf, and on these hills is built the village of Foz; but, by turning this hill, Marius made the waters of the Rhone flow into the gulf of Stomalimne, from which we see that those who made this gulf one of the mouths of the Rhone were not mistaken, as Strabo says that they were" (Statistique du Département des Bouches-du-Rhône) '. But I think Strabo's conclusion is right, for such a canal as that made by Marius did not really constitute a new mouth of the river.

This cut of Marius appears to have been made in a straight line from the Étang de l'Estouma westward to a point on the east branch of the Rhone about a mile above the mouth. The canal was named Fossae Marianae, from which word it is assumed that Foz is derived, the name of a village which stands above the place where the canal entered the salt lake Stomalimne. On one side of the line or hollow, which marks the position of the canal, there has been a cutting into the rock at the base of a hill. West of Foz is the Marais de Foz, which was crossed by the canal. The Marais de Foz terminates in an étang of the same name, which joins the Étang de Galéjon, where it is supposed that in the time of Marius was the outlet of the Massaliotic or eastern branch of the Rhone. The hill which Strabo mentions may be the high ground between Foz and Istres or Distres, which is on the west side of the Étang de Berre. The distance from the

1 This work is by the Comte de Villeneuve, Marseille, tomes i., ii., iii., iv., with an Atlas, 1821-1829. It is a complete description of the department named Bouches-du-Rhône. The archaeological part of this work contains much valuable information, but the author's sketch of the military operations of Marius (ii. 249, &c.) is of no value: it is a piece of fiction founded on a few facts.

commencement of the Fossae (Foz) to the point where it reached the east branch of the Rhone is sixteen Roman miles in the Maritime Itinerary. At this point on the east bank of the Rhone was a port named Gradus Massilitanorum, which may have been established by the Massaliots before the time of Marius. But it is asserted by the author of the Statistique, that the canal of Marius did not receive the water from this part of the Rhone. The canal, he says, was continued due north for about twelve miles from Gradus to the étang of the Desuviates, which comprised the marshes of Arles, of Mont Majour and of the Baux, and into this étang flowed also, at least in part, the Louérion, a canal derived from the river Durance near Orgon. It was the Louérion strictly which supplied the Fossae Marianae. There was a road by land from Massilia, which passed through Calcaria and Fossae to Arelate (Arles). Between Fossae and Arles the road would cross the Campi Lapidei, or the stony plain, now called Crau, part of which is traversed by the railway between Arles and Marseille.

The

Marius made this new cut in order that the Roman vessels might be able to enter the Rhone and bring his supplies up the river into the interior of the Provincia. Without this precaution his army would have been in danger of perishing in a country which had been wasted by the barbarians, and he could not have waited for an opportunity to fight a battle. Marius gave this canal to the city of Massilia for their services to him during the campaign in Gallia. Massaliots derived profit from the canal by exacting tolls from the vessels which passed by it up and down. But the channel, adds Strabo, is still difficult owing to the violence of the stream, the formation of alluvial matter, and the lowness of the country, which cannot be seen from the sea in misty weather even when you are near it. The Massaliots built towers along the coast to serve as beacons, and a temple of the Ephesian Artemis in that part where the outlets of the Rhone form an island.

Though the evidence of the ancient writers as to the great canal of Marius is so small, it is confirmed by an examination of the ground. The authority for what follows is the author