The IP addresses for each of the VMs should be configured in the VM themselves, rather than the host machine.
It helps to picture each of the VMs and the host with their own interfaces, however only the host has a physical cable coming out of it, so the guest VM interfaces need to be bridged to the host interface.
You create the bridge in the /etc/networks/interfaces file, as follows:
auto lo br0 eth0
iface lo inet loopback
iface br0 inet static
bridge_ports eth0
address 192.168.1.10
netmask 255.255.255.0
gateway 192.168.1.1
This will create a new bridge when the network stack is started (such as at boot), and added your interface to it, and given the bridge the IP address of the host. Here you can see the bridge:
$ brctl show
bridge name bridge id STP enabled interfaces
br0 8000.60a4ecf28d84 no eth0
You can treat the br0 interface just as you would the eth0 interface it contains.
Then in the guest config files, you have a line like:
vif = ['bridge=br0, mac=00:16:3E:12:16:19']
This is saying, "give this VM a virtual interface and add it to the br0 bridge, and give it the following mac".
Note that setting a MAC address here isn't needed, but I prefer it so that I can use DHCP to allocate a static IP address to the guests - that way I don't need to hard code any IP addresses other than the host (and the DHCP server, which in my case is a VM itself).
Then in the guest, you just configure it as you would any other linux machine:
auto eth0
iface eth0 inet static
address 192.168.1.11
netmask 255.255.255.0
gateway 192.168.1.1
Note that this is in the guest machine network config.
When you bring the VM up, you'll see that the bridge now has two interfaces in it:
$ brctl show
bridge name bridge id STP enabled interfaces
br0 8000.60a4ecf28d84 no eth0
vif1.0
That vif1.0 is the virtual interface of the guest. Now the guest will be able to ping the gateway and communicate just as if it were directly attached to your network with a bit of cable.