Civil war is a nice card, but its dangerous, especially the first time it is showing up it is damned dangerous. Perhaps making it less disasterous the first time, 5 or 10 tokens should be enough, perhaps even 10 first and 5 the second time, but 5 tokens the first time should be good enough.

The new non-tradables (except regression) is too week. An rise should be good. Perhaps swaping the calamities to new stacks could do the trick, but it would be easy enough to highen them a few unit points or something. More specific ideas is wellcome.

Another thing:

The calamities are of two types, basically: those that do a fixed amount of damage and those that leave a fixed amount of troops left for the player.

Type A:
Volcanic Eruption/Earthquake (lose 5p)
Famine (lose 10p)
Mysticism (lose 9p if you reduce cities on 2s)
Flood* (lose 17p if you live on the flood plain, otherwise less)
Epidemic (lose 16p)
Iconoclasm & Heresy (lose 12p if you reduce cities on 2s)
Piracy (lose 10p and two city sites)

Type B:
Civil War (keep 35p)
Flood* (keep all but flood plain)
Civil Disorder (keep 15 city points)

Type C (kind of in between):
Slave Revolt
Barbarian Hordes

*Where to place Flood depends on the amount of troops on a flood plain.

So type A is a lot more common than type B.

The problem we encountered was Kushan ending his trade with two Civil Disorders... that's a bit cheap when he can just throw the second one. Perhaps there should, for all type B cards, be an extra rule included to cover the situation of two identical cards. Perhaps simply removing an extra city is good enough. Also the point totals lost by the new cards should be counted and compared to the old cards. Personally I like cards that do more damage to specific players (like Barbarian Hordes vs. Africa, Flood vs. Egypt/Babylon) as long as they are tradable.

I guess I'll add more ramblings later on, got to go now...