the same constr. occurs in 74 and Pro Caecina 15, Verr.  Act. I. 24. Antipatrum:  cf. fragm. 1 of Book I. Verbum e verbo:  so 31, D.F. III. 15, T.D. III. 7, not verbum de verbo, which Goer. asserts to be the usual form. Comprehensio:  cf.  I. 41. Ut Graeci:  for the ellipse of the verb cf.  I. 44 ut Democritus. Evidentiam:  other translations proposed by Cic. were illustratio (Quint.  VI. 2, 32) and perspicientia (De Off. I. 15). Fabricemur:  cf. 87, 119, 121. Me appellabat:  Cic. was the great advocate for the Latinisation of Greek terms (D.F. III. 15). Sed tamen:  this often resumes the interrupted narrative, see Madv. Gram. 480. Ipsa evidentia:  note that the verb evidere is not Latin.

Sec.18. Sustinere:  cf. 70. Pertinaciam:  the exact meaning of this may be seen from D.F. II. 107, III. 1.  It denotes the character which cannot recognise a defeat in argument and refuses to see the force of an opponent’s reasoning.  For the application of the term to the Academics, cf. n. on 14, 66, also I. 44 and D.F. V. 94, N.D. I. 13, in the last of which passages the Academy is called procax. Mentitur:  cf. 12. Ita negaret:  this ita corresponds to si below,—­a common sequence of particles in Cic., cf. 19. [Greek:  Akatalepton]:  the conj. of Turnebus [Greek:  katalepton] is unnecessary, on account of the negative contained in negaret. Visum:  cf.  I. 40. Trivimus:  cf.  I. 27. Visum igitur:  the Greek of this definition will be found in Zeller 86.  The words impressum effictumque are equivalent to [Greek:  enapesphragismene kai enapomemagmene] in the Gk.  It must not be forgotten that the Stoics held a sensation to be a real alteration ([Greek:  heteroiosis]) of the material substance of the soul through the action of some external thing, which impresses its image on the soul as a seal does on wax, cf.  Zeller 76 and 77 with footnotes. Ex eo unde esset ... unde non esset:  this translation corresponds closely to the definition given by Sextus in four out of the six passages referred to by Zeller (in Adv.  Math. VIII. 86 Pyrrh.  Hypotyp. III. 242, the definition is clipt), and in Diog.  Laert.  VII. 50 (in 46 he gives a clipt form like that of Sextus in the two passages just referred to).  It is worth remarking (as Petrus Valentia did, p. 290 of Orelli’s reprint of his Academica) that Cic. omits to represent the words [Greek:  kat’ auto to hyparchon].  Sextus Adv.  Math. VII. 249 considers them essential to the definition and instances Orestes who looking at Electra, mistook her for an Erinys.  The [Greek:  phantasia] therefore which he had although [Greek:  apo hyparchontos] (proceeding from an actually existent thing) was not [Greek:  kata to hyparchon], i.e. did not