Bury—On the Roman Empire in Fourth Century, A.D. 493 
does not tell us who succeeded them under Julian’s régime, but as 
Mamertinus was praetorian prefect in 362, we may suppose that he was 
their direct successor, created perhaps at the same time as Secundus 
Sallustius, or shortly after. In the first months of Julian’s reign we 
find him empanelled as assessor to Sallustius, quaestionum agitandarum, 
and perhaps this implies that he was not yet prefect: xxir., 3, 1, ‘‘ Brevi 
deinde Secundo Sallustio promoto praefecto praetorio summam quaes- 
tionum agitandarum ut fido conmisit: Mamertino et Arbitione et 
Agilone atque Nevitta adjunctis itidemque Jovino,” &c. This com- 
mission dealt with the officials of Constantius, and one of their acts 
was to send Taurus into exile. 
This fact, that after the accession of Julian, Mamertinus became 
sole prefect of Italy and Illyricum, which had in the immediately pre- 
ceding years been governed by two prefects, proves that no definite 
administrative arrangement had been made fixing or limiting the 
number of praetorian prefects ; and as a consequence of this there was 
no definite division of the empire into prefectures except in so far as 
custom prescribed. I may add that the territorial division followed 
the number of the officials, and not vice versa. Valentinian and 
Valens did not primarily divide the land—they divided the palatium. 
Having thus established that Zosimos’ statement is not correct, we 
may proceed to examine the mode in which the multiplication of prae- 
torian prefects, and their assignation to particular parts of the empire 
arose. 
Diocletian’s division of the empire is generally called a quadri- 
partition, but it was essentially a bipartition. Diocletian was 
Augustus of the whole Eastern half, including the dioceses of the 
Pannoniae and the Moesiae; under him the Caesar Galerius had a 
delegated sway over a certain part. Similarly Maximian was Augustus 
of the West, and Constantius his representative in Gaul. 
Praxagoras, who lived in the early part of the fourth century, and 
wrote a history of Constantine the Great, gives the arrangement of 
territory thus (Miller, F. H. G., rv. p. 2):—6 zarjp Kwvorartivov 
Kovoravtios Bpetavias é€Bacidevoe, Magipivos (Maximianus) d€ tis 
Popys kat THs GAAns IraXdias kai Suxedias, 6 d€ erepos Maéiptvos 
(Galerius Maximianus) rijs te “EAAdOos kal Tis Kato “Acias kal Opaxns. 
AvoxAntiaves 6€ 6 Kai Tav GAAwy tpecBitatos THS Te Buibvvias HpxXe Kat 
7s ApaBias kat THs AuBvns kai THs Aiydmtov ony 6 NeiXdos érepxopevos 
apdec. Macedonia, Dacia, and Pannonia are not mentioned, but they, 
doubtless, went with Hellas and Thrace. 
But we must not be misled into thinking that this was equivalent 
to a quadripartition of the Empire. For Constantius and Galerius 
were merely helpmates of the Augusti—merely governors of very large 
provinces, with the prospect, however, of becoming Augusti them- 
selves at some future time. There was no strict division between the 
provinces ruled by Diocletian and Galerius, such as there was between 
the dominions of the two Augusti. Diocletian was as constantly 
in the Balkan peninsula as in Egypt or Asia, and Maximian was 
