Help?

Guys,

I've been studying for an exam and came across a question which I'm stuck on and I'm wondering if any of yous can help me out. It's as below:

When looking at data packets on a network, how do you know which host is the client and which is the server?

1. Client will request a destination port in the dynamic range and the server will be communicating with destination port in the common/application range.
2. The client will request a low sequence number. The server will be communicating on a higher sequence number
3. Client will request a destination port in the common/application and the server will be communicating with destination port in the dynamic range
4. The client will request a destination IP greater than it's own. The server will be communicating on an IP in the 200's.

Any help would be appreciated guys.