excess of salt from the water, so that the particles of the sand cohere and ultimately become stones.
20. the porous stone which changes to pumice when it is fired.
Both the name and the nature of this stone are uncertain. Wimmer retains the manuscript reading διαβάρον, which is accepted by all the editors except Hill, who substituted the word 'Αραβικού (printed incorrectly with a rough breathing) and translated it as "Arabic stone," which according to Dioscorides94 and Pliny95 is a substance resembling ivory. There seems to be little justification for this conjecture. Since Theophrastus is discussing pumice and similar cellular stones, it seems probable that there is merely a slight error in spelling and that the reading should be 8ιαβόρον, a word meaning "porous." The stone itself may have been a volcanic tufa of some kind.
The negative οϋ, which occurs before κισσ-ηρονται. in the Aldine edition but is not in the manuscripts, is a difficult reading to interpret. If it is accepted, the passage refers to a stone which "does not change to pumice when it is fired." However, the οΰ in the Aldine edition is in the form of an abbreviation of the same size as the relative pronoun η ("which"); this occurs in manuscripts Β and C, though ή is written with a smooth breathing in manuscript A. Most of the editors prefer ή και (i.e., "which also changes"), but Wimmer accepts the reading of Aldus. Since this is the more difficult reading and therefore harder to explain, it is easy to see why Wimmer felt obliged to adopt it. Actually, the evidence of the manuscripts seems to carry more weight, and the acceptance of the reading ή would make it easier to understand the meaning of the passage. There are many misprints in the Aldine edition; it is quite possible that the abbreviation representing ov is not correct, and there is no evidence that Aldus had another manuscript from which he might have obtained this reading. It seems best to accept η, since it appears in two of the manuscripts. If η ("or"), which appears in A, is correct, then some other verb must have dropped out before it; but it is possible that the smooth breathing is a mistake, and certainly A provides no evidence of a
»*V, 148 (Wellmann ed., V, 131). 95 XXXVI, 153.
