Without seeing it, hard to say. Chances are that it's condensed elemental mercury though. You will also occasionally see this if a lamp has been chilled for an extended period then run up. The mercury will natrually consense in the coldest location.
I've seen that phenomenon to some extent in new linear fluorescents, though not in CFL's that I can remember. If it disappears, that's probably what it was. If not, possibly a defect in the phosphor coating.
Think this is another point where photos can speak a thousand words though.
