From what I understand, each time openGL draws a frame, it culls whatever was there last time. So each frame you need to specify the location/transformation of each object + camera + lights.
In order to display more than 1 object, you'll need a way of storing information about this object that can be used when drawing it.
So, each time you create an object with the mouse, save all the information needed to draw it & add this information to a list(that you create and maintain) of all objects in the scene.
Then when it comes time to draw the scene, just step through the list of created objects, transforming & drawing each one as you go.
simple openGlut pseudocode may resemble:
vector pos1, pos2
-- do the steps to draw object1
do the steps to draw object2
I'm only _very_ new to this, having first tried my hand at openGL earlier this week. But this is the understanding that I (think?!) I have.