technician and mark.watson both are on the right track, but I would say technician is assuming a constant force, while mark.watson is giving an infinitesimal form of the actual definition:
The definition of impulse I is
\vec{I}=\int \vec{F} \ dt Impulse is a time integral of force.
Momentum, on the other hand, is a property of an object. If an impulse I is imparted on an object, it will change that object's momentum by \vec{I}. To be more explicit, if the object initially has momentum \vec{p}, and then an impulse \vec{I} acts on that object, its final momentum will be \vec{p}+\vec{I}. So an impulse is basically momentum transferred.