Zosimus sources on army strengths

[Aryaman2]

Hi all
This list is taken from one provided by Vortigern in a previous thread, I have listed by date
Late Roman Army strenghts:

188.000 - army of Maxentius, 312 AD (Zosimus II.15.2)
98.000 - army of Constantine, 312 (Zosimus II.15.1)
165.000 - army of Licinius, 324 (Zosimus II.22.1)
130.000 - army of Constantine, 324 (Zosimus II.15.2)
130.000 - army of Licinius at Chrysopolis, 324 (Zosimus II.26)
65.000 - main Roman army in Persia, 363 (Zosimus III.13.1)
23.000 - army of Julian, 360 (Zosimus III.10.2)
18.000 - army of Procopius, Persia 363 (Zosimus III.12.5)

The forces listed can be grouped in 2 periods, that of Constantine and that of Julian. Now, it is clear that the armies of Constantine period are much larger, what I wonder if is that could be explained by itself, or would it be a consequence of less reliable sources available to Zosimus for an earlier period?

[Mitra]

Probably the numbers for Costantine, Maxentius in 312 are the total men in the rolls, not the field army.

For the 324 i agree with Treagold, Zosimus can have followed a strange logic: for Costantine the roll's army of Balcans taken from Licinius recently,for Licinius the rolls of ex-Maximinus Daia army.

If we make how Treagold make, a parallel calculation on 235 and notitia dignitatum numbers (adding and subtracting the changing in the century), we have +- 130000 men for the Balcans and 165000 for Egypt and Asia.

[Robert Vermaat]

I don't agree with Davide, Zosimus is talking about the actual numbers of troops for the campaign. But that does not mean a very big Roman army, it means field army plus a whole lot of hired foreign troops, which were disbanded after the campaign was over. That, too, is mentioned in the sources.

No, I disagree with the notion that a high number of troops should mean unreliable sources.