Internet email is highly dynamic. Hosts can be up/down. Routers can be up/down. Gateways can be up/down. Routing databases can be out of date. Cables can be cut. Load averages can affect whether a host accepts SMTP connections. And so on. Sending one message and rigorously studying timestamps in the Received lines and using that to draw conclusions about mail propagation is a complete waste of time. /r$