22 PETRONIUS ARBITER 
In several of these classes, especially the fourth, traces of a regular series, although 
with some exceptions, are perceptible. If we unite the last five classes — the third, 
fourth, fifth, sixth, and seventh —into one group, calling it the linguistic, dnd the first 
two into another, the realistic, we have two nearly complete series, the regularity of 
which is disturbed by two expressions only, viz.: 31, Intra Kalendas, and 50, Dipsas, 
which belong to the first instead of the second series. Whatever importance may attach 
to these traces of a regular arrangement, it is certain that the compiler aimed neither at 
perfect regularity nor completeness. Even a glance shows that there is no attempt at 
an alphabetical arrangement. : 
As to the age of the compiler, a late period must undoubtedly be assigned to him. 
Both the language and contents of some of the articles point to Christian sources, and 
if these are the writings of such men as Hieronymus and Augustinus, as appears prob- 
able, the author of our fragment, whether his name be Petronius or not, cannot be 
placed earlier than the end of the fourth or beginning of the fifth century of the Chris- 
tian era. 
Not the least interesting question, however, relates to the sources of our author, and 
especially his relation to Gellius furnishes a curious problem. If we examine the 
arrangement of those words which occur in Gellius as well as Petronius, they form, 
with one exception, an uninterrupted series of twenty-three words, viz.: 20, Avarus; 
21, Locuples; 22, Lex; 23, Populus; 24, Plebs; 25, Plebiscitum ; 26, Priya; 28, 
Elegantia; 29, Mendacium ; 30, Lictorem; 31, Intra Kalendas; 32, Pomerium ; 99, 
Humanitas; 34, Manubiae ; 35, Faciem ; 36, Profligare; 37, Vestibulum ; 38, Vescum ; 
39; Foenerator ; 40, Assiduus ; 41, Quin; 42, Soror; 43, Frater. The exception referred 
to is the article 27, Italia, which intervenes between 26, Priva, and 28, Elegantia. It 
will not be denied that this circumstance is by itself a strong, although not irresist- 
ible, argument in favor of the opinion, that Petronius borrowed in all these articles 
directly from Gellius; that while reading the work of Gellius he copied, with more 
or less completeness and accuracy, the definitions of Gellius. This view is strength- 
ened by another circumstance, namely, that Petronius, with two exceptions, follows in 
these twenty-three articles the order of Gellius, the two exceptions being the articles 
42, Soror, and 43, Frater, which in Gellius occur in lib. 13, 10, earlier, therefore, than 
the ten preceding articles, from 32, Pomerium, to 41, Quin. Another less important 
departure from the order of Gellius may be mentioned, viz. 40, Assiduus, which i in 
- Gellius precedes, while in Petronius it follows, the article 39, Foenerator. . 
On the other hand, it should not be overlooked that in some of these articles, a as has | 
been pointed out, the agreement between Petronius and Gellius is less striking, and 
