I'm 100% certain of the answer to your question, but again it appears to me that your author is mixing up her logic a little: the tone falls on the morpheme: the morpheme doesn't fall on an unstressed syllable - it is a component of the syllable.

Literal translation:

"In the phonological process of tone deletion, contour tones behave as indivisible units. For example, in word FORMATION (a, b) and syllablE reduction processes (c, d), contour tones of morphemes that do not occur in unstressed syllables are totally deleted/deleted in their entirety."

Another possibility:

"In the phonological process of tone deletion, contour tones behave as indivisible units. For example, in word FORMATION (a, b) and syllablE reduction processes (c, d), contour tones that do not fall on unstressed syllables are totally deleted/deleted in their entirety."

Or "contour tones not on morphemes in unstressed syllables.."

In any case (without quoting me), unless this makes good sense to you, I'm wondering if you might want to double-check what your author means, because she could be trying to say the opposite: that tones on unstressed syllables are deleted (which would make a lot more sense to me, though I don't know much about tone languages).

--------------------------------------------------
Note added at 21 days (2007-08-27 09:21:12 GMT) Post-grading
--------------------------------------------------

Then she had an extra negative in her statement. Based on what she has now said, the translation would be the oppostie - that is:

"In the phonological process of tone deletion, contour tones behave as indivisible units. For example, in word FORMATION (a, b) and syllablE reduction processes (c, d), contour tones of morphemes in unstressed syllables are deleted in their entirety."