SQL & PL/SQL :: Conditional Group By Of Rows

Task I would like to do in plain SQL rather than PL/SQL.

Given this table:

create table TEMP_TEST
(
DT_START DATE not null,
DT_END DATE not null,
FIELD VARCHAR2(2)
)

and these rows:

insert into temp_test (DT_START, DT_END, FIELD)
values (to_date('01-01-2005', 'dd-mm-yyyy'), to_date('31-08-2007', 'dd-mm-yyyy'), 'ML');
insert into temp_test (DT_START, DT_END, FIELD)
values (to_date('01-09-2007', 'dd-mm-yyyy'), to_date('31-12-2007', 'dd-mm-yyyy'), 'MT');

[Code]....

what i want to do is collapse the rows that have same field value, ONLY if the end date of the first one equals start date -1 of the next one.

that being said the result should be

01/01/2005 - 31/08/2007 - ML 01/09/2007 - 31/12/2007 - MT 01/01/2008 - 29/02/2012 - ML 15/03/2012 - 31/12/9999 - ML

I was doing this:

SELECT MIN(dt_start), MAX(dt_end), field
FROM(
SELECT dt_start,
dt_end,
field,

[Code]....

but it still misses the thing about adjacent time periods