To understand how to subnet, you have to step back and think a little about how routers make routing decisions and how they relate to the subnet mask. Further, you have to understand what composes an IP address and how a subnet mask divides the address up.

Let's take a typical Class C address network like 223.1.2.0 and break it down a little. All IP addresses are composed of 4 byte values or octets. The binary form of this address would look like:

11011111 . 00000001 . 00000010 . 00000000

By applying a mask of 255.255.255.0 or:

|        Network Number     |   | node |
========================================
11111111 . 11111111. 11111111 . 00000000

Below you will see all possible ways to break up a Class C address. Each scenario yields a different number of subnetworks and nodes in each subnetwork depending on the subnet mask you select. Since each subnet mask must break on a bit boundry, there are 8 possible masks. The last case is really useless since it yields no node space, only a network number and broadcast address for each network. Click the Show Table links to generate a table suitable to help plan a subnetting scheme. You do need a broswer that supports tables though. The pages are usefull to print out and help document how you subnet your networks and possibly to help configure clients and log their IP's.