Is spatial containment really a transitive relation?

I think containment within containment is a mathematical abstraction. For example: 

"The water is contained in the bottle and the bottle is in the room. Therefore the water is in the room."

You want to comment that for the water to be "in" the room could mean one of several things - in the bottle, as a spray in the air, as a puddle on the floor, etc. In each case you have to stop and think: yeah that is "in" the room. Because "in" is an abstraction covering these kinds of diversity.

Update: But even without appeals to general properties of being "in", language does not support the transitivity of water in a bottle also being in a room. The bottle is in the room, Period. Everything else is a technicality not an intuition. To think of the water as being "in" the room requires removing the bottle.  

Update: I am going to keep worrying at this. I think containment is a natural intuition but "in" is an abstraction that is defined to include contained containment. You say the water is contained in the bottle but you do not say the water is contained in the room. So really, since you can define "in" as a sequence of containments, what is wrong with that?