My thanks to Andrew, Michael, & Tony, who replied to my earlier
misguided email on this thread. Your suggestions were very helpful to me.

I eventually tracked down the cause of the problem in terms of the final
output in my browser, which was my use of the Zend Search Lucene
$query->highlightMatches() method, which wraps the returned data in html
tags, and which upon seeing a 2-byte character such as a UTF-8 Spanish
n: q, etc., converts each of those two bytes into separate html
entities, which the browser then displays as: C1.

My apologies also for the mis-naming of this thread, which as Andrew
pointed out, should be "character encoding" rather than "character
entities".