Purser—On a London MS. of Cicero’s Letters. 369 
latus. We find H at one with Erf. in the following :—359. 2, 
intellegentia; 7, naturarum; 19, omnibus; 360. 9, atque (for 
quod); 23, caupadoces (H has above the line uel caupones; 29, r (for 
equites Romani); 82, optimusque caucilium; 361. 10, curiose tan- 
nios; 362. 8, ab honoribus; 15, petitionem magistratus; 81, con- 
sul; 3863. 4, c. fundanique gali chociuii (Erf. has chorcinii); 364. 
5, inter nos calumniatores ; 29, hominem quam iners; 365. 27, autem 
emi quod; 31, obediendo; 366. 17, adspectatorem ; 367. 31, facete 
abs; 368. 1, equandum dato; 3869. 34, y (for Romani). The 
(ufferences are slight, and such as would be made by two different 
copyists of the same archetype. H is rather more correct than 
Erf., which latter Wunder thinks of no very great importance 
in this treatise; ¢. g. H has: 359. 7, descendenti; 19, fere; 360. 
10, illis; 15, homini; 362. 34, prorsus; 363. 4, deferundis; 6, 
est wserted; 17, homines inserted; 364. 1, hominum; 9, compa- 
rantur; 365. 28, adhibebitur; 36, salutatorum; 366. 26, honesta- 
tem; 367. 13, melius; 38, diurni nocturnique; 369. 21, ornando ; 
28, poscit. 
The Erf. ms. contains the De Amicitia and the De Senectute; but I 
have been unable to find such resemblance as would lead us to attribute 
the copies in H to the same archetype. As regards the De Amicitia, 
I compared the first fifty variants (of Ernesti’s edition from Erf.), as 
given by Wunder, and found twenty-seven agreements of H and Erf. 
In all these cases, except two (viz. § 1, augur sceuola; § 9, nec 
catoni comparantur, the two mss. preserve what is really the more 
correct reading, erroneously altered by Ernesti; and in the remaining 
twenty-three, H has the correct reading, which has been corrupted 
by Erf. It has nec sicut . . . sapientem (§ 7), which is omitted by 
Erf., and, also, et uere (tuere, Erf.) in § 8. Of the first fifty impor- 
tant variants of the De Senectute, H agrees with Erf. in twenty-seven. 
Of these it agrees five times wrongly, viz., § 1, flaminium, attice ; 
§ 2, leuare, certe scio (see Mr. Reid’s note ad loc.); § 8, ignobilis. 
Out of the twenty-three times H disagrees with Erf., it does so only 
twice wrongly : § 5, ferendum; § 6, ingrediendum. H. has all the 
words in the first ten sections, ‘which Wunder notices as omitted by 
Erf. In § 3, it has attribuito corrected into attribuitur, and, in 
§ 10, a mixture of two readings, viz., cum etate condita erauitas 
cum etate condita uirtus grauis. H is on the whole wonderfully 
well copied. There is very considerable similarity between H and 
what Graevius calls his primus; but they are not the same, nor, 
I think, of the same family; for though several examples of agree- 
ment may be adduced, especially in the first ten sections or so, 
still, in the latter part of the treatise, such variants as 600. 26 
(Orelli), occatum (occaecatum Gr.); 33, oblectamentumque (oblecta- 
menta Gr.); 608. 25, inclusi (conclusi Gr.); 609. 28, creditote 
(eredite Gir.), 610. 4, colitote (colite Gr.); 23, stultissimus ini- 
quissimo (stultissimus aequo, G7.), must be considered of great 
