Today I outlined the arguments, pro and con, relating to the idea (very old now; I also found some old references on the subject) that bulges form through the mergers of disks. I read an old paper by Barnes that showed the results of a simulation in which two disks merge to form something that looked like an old bulge, and the bizarre thing is that the bulge so created had a higher central density than either of the two original disks. This goes against common sense, and against the principle that a merger heats the remnant relative to the inputs (the entire motivation for the study in the first place: disks cold, bulges hot). I will email Barnes in the morning for an explanation. I wonder what the current state of the art is.