In a message dated 5/14/99 1:02:31 PM Pacific Daylight Time,
[log in to unmask] writes:

<< The thing that I have found most disconcerting is that there is
 virtually no empirical evidence for (3).  But without (3) natural
 selection (items (1) and (2)) can only interact with the
 variation that is already present in the gene pool.  Since
 biologists are unhappy with this, (3) is *assumed* to be true. >>

I don't think it is a question of being unhappy but that a mutation is, by
definition, a random event.