On the issue of separate subnets, I'm aware of a couple of local schools that have multiple subnets although I can't see the benefit myself. Historically I believe subnets were used to reduce the amount of broadcast traffic seen by each PC (because the subnet is effectively a boundary to broadcast traffic - a broadcast domain) and to increase security by restricting traffic flow (because your routers/layer 3 switches act as passport control between subnets).
Especially with a school as small as yours the amount of broadcast traffic will typically be negligible. We have a lot more kit on a single subnet and the broadcast traffic is almost nothing. Even if we had a split site with a poxy 10mb link it wouldn't be significant.
As far as increasing security - others have said it - there are plenty of tools at your disposal in a standard Windows network to restrict access to the things schools typically worry about. Any gain you might get in security could easily be turned into a loss by the increased administrative burden you are put under, meaning you have little or no time to monitor, maintain and update your security measures.
Think of the numerous scenarios where staff might want to access the same resource from both networks. Think of the hassle of them having to pick the right PC to perform a particular job (in general that is what we are trying to move away from - surely it should be the person's identity that determines their access rather than their location)
I have taken part in network merges at two large schools and I believe staff at both would say the merge led to a much smoother ICT experience.