Are you asking rhetorically, or do you not know?

For those that are playing along at home, the point of series termination resistors is to match the driver impedance to the trace impedance. Assuming an unterminated far end, the reflection from the far end gets completely absorbed in the termination and so doesn't bounce back again off the driver. There's more to it than this very simple explanation, but that's the crux of the biscuit.

One could use parallel, or Thevenin, termination at the receiver (a-la ancient VME recommendations, etc) but there's a serious power cost when you have sixty-four 500-ohm resistors hanging across a power rail.

A hack you sometimes see is the small cap (10 pF or less) to ground at the receiver end of the line. Sometimes a large resistor (1 M) is put in parallel with the cap. (This is called "ship it with a 'scope probe attached" because the circuit magically works when being probed.) This works with signals that have near 50% duty cycle (like clocks) but isn't very good for "irregular" signals as it affects the rise and fall times (you are, after all, charging a capacitor) and some inputs may not be happy with slower transitions near the logic thresholds.

I highly recommend reading the books by Johnson and Graham, High-Speed Digital Design: A Handbook of Black Magic, and High-Speed Signal Propagation: Advanced Black Magic. The usual effects on engineers reading these books include wanting to go back and re-design every PCB!

--a