South of the Charadra the plain is highly cultivated, mainly with the vine. If it was so in 490 B.C., it would not have been by any means unimpeded ground for cavalry, as the Greek habit of allowing the vine-tendrils to run along the surface renders the vineyards of the country troublesome, if not dangerous, obstacles to the passage of a mounted man going at more than a foot’s pace. North of the Charadra the cultivation is more sparse and the ground more open. From the plain two roads lead to Athens: (1) An upper one, which leaves the plain in two branches, one of which goes up the valley of the Charadra past the modern Marathona, and then bends W.S.W. through the hills in the direction of Athens. About a mile and a half beyond Marathona it is joined by the second branch, which, leaving the plain near the modern Vrana, goes up the valley of Avlona and through the site of the precinct of Herakles, which is situated near, though not at the point of junction of the two. From this point the road proceeds over the mountains, and, passing the modern Stamata, descends into the Athenian plain near Kephisia, whence it goes direct to Athens. It is a rugged track, though not very difficult, and is certainly that which was used by the Athenians in their advance to Marathon. [Illustration: THE “SOROS” AT MARATHON. [_To face page_ 165. ] (2) A lower road, practically coincident with the modern carriage road from Marathon to Athens. It is much easier than the upper one. The country it traverses is easy. There are no steep gradients. It passes through a long stretch of thin pine forest. It leaves the plain by the narrow passage between the little marsh and Mount Agrieliki. The important feature of this road is that it is, and must always have been, easily traversable by cavalry. I mention this, because some emphatic statements to the contrary have been made in some modern treatises on the subject of the battle. The only two landmarks on or near the field which can be said to be in any way decisive as to the history of the battle are both artificial in origin. They are— (1) The precinct of Herakles, already mentioned. The site was determined by Lolling in 1876.[66] (2) The soros, or mound, raised over the Athenian dead. The absence of any mention of this by Herodotus is of itself significant evidence that he had never visited the ground. It stands, as will be seen from the map, about half a mile from the shore, and about a mile and a quarter south of the Charadra. Pausanias (i. 32. 3–5,7) mentions it in his description of the neighbourhood. That description bears all the traces of being founded on autopsy. Of the grave of the Athenians he says, “There is a grave of the Athenians in the plain, and on it are columns with the names of the dead upon them, according to their tribes.”