How to eliminate duplicate accounts selectively from a SQL Server 2005 table?

I am developing an Access application usng Access 2003,

I use an ADP type file with SQL Server 2005.

I have a table named tblGetDocumentAttributes with fileds including the following:

      MailID   Account   DocumentDate                     ScanDate    Amount.
 1   A34R   45677      2010-08-31 00:00:00.000
 2   B78E    45677      2010-09-30 00:00:00.000
 3   R546    45677      2011-01-31 00:00:00.000
 4   G555    45677      2011-01-31 00:00:00.000
 5   C645    86453      2010-08-31 00:00:00.000
 6   E567    86453      2010-09-30 00:00:00.000
 7   R546    86543      2011-01-31 00:00:00.000
 8   R555    86543      2011-02-28 00:00:00.000


Can you think of a SQL statement to delete any records from this table in which
any particular account has multiple records with the same documentdate and in these instances just keep the record with the higher MailID (type nvarchar) ?

So in the example above, the 4th record would be deleted because it has the same
account plus documentdate as record 3 but record 4 has the lower ranking MailID.