We discuss the situation where attractive and repulsive portions of the int
erparticle potential both contribute significantly to glass formation. We i
ntroduce the square-well potential as prototypical model for this situation
, and reject the Baxter model as a useful model for comparison to experimen
t on glasses, based on our treatment within mode coupling theory. We presen
t explicit results for various well widths, and show that, for narrow wells
, there is a useful analytical formula that would be suitable for experimen
talists working in the field of colloidal science. We raise the question as
to whether, in a more exact treatment, the sticky-sphere limit might have
an infinite glass transition temperature or a high but finite one.