I've been puzzling over the following thought experiment: Consider two versions of a perfectly rigid, sealable box. Suppose first we fill the first box with water at sea level, so the hydrostatic pressure inside the box is basically 1 atm, plus the pressure of the water (but let's say the box is very shallow). Now seal the box and push down into a deep body of water and attach it there.
Second, suppose that we leave the second box open, attach it under water, and then seal it there. In that case, the pressure inside the box is the same as the (high) ambient water pressure.
How should I imagine the difference between the two boxes? The first box experiences strong inward pressure (driving it to implode), which we assume is absorbed by the box, whereas the second one does not. (Only the material of the box itself is under pressure.) But the water inside and outside is incompressible, and there's nowhere for the force to act, no energy to spend (?).
Is there any theoretical mechanism by which one state could be transformed into the other, in either direction? It seems that displacing content is not an option. What if I slowly open the seal of the first box -- does the interior pressure immediately go up? Is there any net force on the lid?
(Aspects of this question appear in this related question about valves.)
Box 1:
+-------+
| Water | ---->
--+-------+--
\~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~
.
.
.
.
+-------+
| Water |
+-------+
##########################
Box 2:
/
/
/
+
| Air | ---->
--+-------+--
\~~~~~~~ ~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~
. .
. .
. .
---->
. .
. .
. .
/
/
/
+ +-------+
| | | Water |
+-------+ +--------
##################### #################