To separate the API from the implementation, which is often a good programming practice. It will help with readability, if nothing else. It will also allow someone using your code in the future to provide an alternate implementation of the interface if they so desire.