cato (De agricultura 18–19) in the 2nd century bc provided the first detailed description of a press room. He describes a beam or lever press. This would be constructed on an elevated concrete platform with a raised curb, which formed a shallow basin, which sloped gently to a run-off point. On this was constructed the press, which consisted of a long, heavy horizontal beam, which slotted into an upright at the back and ran between two uprights at the front. The front end of the beam was attached by a rope to a windlass. The grape solids were put under the beam and pressure applied by winding down the end. As the pulp compacted, so wedges were hammered into the slot at the pivot end to lower it. Over time various refinements were introduced. Most notably, according to pliny (Natural History 18. 317), a ‘Greek-style’ press was introduced in the late republic or early empire, in which the windlass was replaced with a vertical screw thread, sometimes with a heavy counterweight. There is ample archaeological evidence from Italy, and elsewhere, for the use of presses. All the Roman agricultural treatises, apart from the writings of Palladius, assume the use of a press in their descriptions of winemaking. However, the press was an elaborate and comparatively expensive piece of equipment and its use was far from universal. Some farmsteads have large tanks for treading the grapes in, but no evidence of a press. It is not clear whether the must from the treading was always kept separate from that from the pressing. The grape pulp could be subject to a second pressing; but this was carefully kept separate. The pressed grape skins could even be soaked in water to produce lorca, a drink to be given to the farm hands (varro, De re rustica 1. 54), a forerunner of piquette.