Dense_Rank Revisited

[Pete Helgren <pete@xxxxxxxxxx>]

It looked like dense_rank did the trick with my sample data but when I 
had a larger data set I ran into a snag that may not have been apparent 
in the example I presented.  Here is a slightly different variation that 
describes the remaining issue.

BB8179      04/17/11     04/02/11
BE7214      04/17/11     04/02/11
BE7488      04/18/11     04/02/11
BE2178      04/18/11     04/02/11
BE1618      04/18/11     04/02/11
BD9608      04/21/11     04/02/11
BE2180      04/21/11     04/02/11
BE1696      04/21/11     04/02/11
BD9607      05/7/11      05/28/11
BB6382      05/7/11      05/28/11
BB7942      05/10/11     05/28/11
BE7487      05/10/11     05/28/11
BE7489      05/11/11     05/28/11
BE2179      05/11/11     05/28/11
BE8955      05/11/11     05/28/11


What I would like to do is count the unique occurrences of the first
date column within the *same* occurrence of the second date column.
Something like:

BB8179      04/17/11  1   04/02/11
BE7214      04/17/11  1   04/02/11
BE7488      04/18/11  2   04/02/11
BE2178      04/18/11  2   04/02/11
BE1618      04/18/11  2   04/02/11
BD9608      04/21/11  3   04/02/11
BE2180      04/21/11  3   04/02/11
BE1696      04/21/11  3   04/02/11
BD9607      05/7/11   1   05/28/11
BB6382      05/7/11   1   05/28/11
BB7942      05/10/11  2   05/28/11
BE7487      05/10/11  2   05/28/11
BE7489      05/11/11  3   05/28/11
BE2179      05/11/11  3   05/28/11
BE8955      05/11/11  3   05/28/11


I tried adding the second column to the order by clause but that seemed 
to make no difference.  I also tried 'partition' by using the second 
column but that also didn't deliver the desired result.  Ideas?