BOOK XXV. Li. 91-L111. 93 



LI. Dogs too have found a plant by which they Piants found 

 cure loss of appetite, and eat it in our sight, but in ^ "''*■ 

 such a way that it can never be identified, for it is 

 seen only when chewed up. This animal shows yet 

 greater spitefulness in its secrecy about another 

 plant ; for there is one by which it is said to cure itself 

 when bitten by a snake, but it does not crop it when 

 a human being is looking on. 



LII. With greater frankness deer have shown us Pianisfuund 

 elaphoboscon, about which we have written," and ^ "'"' 

 after yeaning have made known seseHs and the black 

 bryony, as we have pointed out ^ ; (LIII.) dittany also 

 by feeding on it when wounded, the weapons at once 

 falling out.'' The latter grows nowhere except in 

 Crete, with branches very slender; it resembles 

 pennyroyal and is burning and harsh to the taste. 

 Only the leaves are employed ; it has no flower, no 

 seed and no stem ; its root is slender and without 

 medicinal value. Even in Crete it does not grow 

 widely, and the goats are wonderfuUy eager to hunt it 

 out. A substitute for it is false dittany, which 

 grows in many lands, like true dittany in leaf but 1 



with smaller branches, and called by some chondris. 



vere percussi eo telo pastuque herbae eius eieeto. A scribe 

 with an aural memory of XXII. § 41 may have had it suggested 

 to him by tlie similarity of dictam to et tamnum with diciamnum 

 foUowing a httle later. Seselin is sufficiently like helxinen to 

 the eye to set in motion latent aural memory, and the other 

 change, the omission of a connecting particle before dictamnum, 

 would follow naturaUy. Hermolaus Barbarus in 1492 saw 

 that some restoration of the text was necessary, and would 

 delete dictammim before imlneratae and emend dictam to 

 dictamnum. Whatever the details of the correct emendation 

 may be, the corruption took place very early, before the 

 archetype of our existing MSS. was written, as none of them has 

 any sign that helxine(n) and dictam are wrong. 



203 



