excellent thinking. I agree... we need to start thinking more specifically and stop pulling crazy patterns out of the sky... we're just getting ourselves confused here...

I think the idea that each dial has an independent solution has merit- there is much more order to breaking the code into self-contained units rather than spreading it across the board without any hint of an organizing structure. I want to repost Chelsea's latest clue here to maybe help guide along some thinking.

"The hands of the clock ticked strangely--almost as strangely as the clock was numbered. What was worse, she couldn't tell what was supposed to be there, and what wasn't.

"Is there something missing? Something not? Something else?" she asked herself. She saw the patterns, she thought, but couldn't be sure.

And the birds kept calling. "

First off, I think (and somebody esle already posted something like this) that "the birds" refer to the solution... after all.. the "birds falling" post came when we were beginning to knock down the central gylphs... At this point, "birds" seems to amorphous to represent any quantity or letter (i could be wrong though...)

The clocks therefore are our biggest clue. If each dial is indeed independent, chelsea's clue may give us a method (albeit vague) we can use to solve them. When she states "the hands of the clock ticked strangely" it implies there is a type of progression among the numbers... and that its not a standard progression. What's more, she implies that there is something missing.

The logic process of the "she" in the post is a good one: 1) find out what is missing, 2) find out what is not missing, 3) relate the findings to "something else" (most likely a letter). The language of the clue implies that there is a near-pattern in the dials... and that to complete it we must *add* numbers... the "find what's not missing" is a little ambiguous... i'm not sure if it means remove extraneous information or simply just recognize what fits....

If we can figure out the patterns (and there very possibly is more than one pattern used... maybe even separate patterns for each dial) then finding the U/P should be obvious. Perhaps each dial will produce two letters, or maybe the message formed will have some obvious separation in it. Until we find the patterns, though, we shouldn't guess. Working on what we think will work will run us to a dead end... lets just concentrate on finding a pattern *then* analyzing it.