Dear Mike, I have tried to demonstrate that the inequality you yourself came up with amounts to the same thing, but for the "negative" cases of particles moving right to left. This is nothing more than a convention (in the same way that one has to be careful to keep to convention when considering ray optics). The labelling of the particles is totally arbitrary and either way round you care to look at it, you are left with an inequality which cannot be supported. Furthermore, the physics argument is perfectly straightforward. If some kinetic energy is lost in an inelastic collision, then either some mass or some velocity, or both, has been lost. But if it's been lost, then some linear momentum (which is itself a product of nothing more than mass and velocity) has also been lost. It doesn't just pop back in again, so as to save a so-called "law." Neville. Mike <mboyd@xxxxxxxxxxxx> wrote: Dear Neville, The fact is that if you divide an inequality equation by -1 you change the sense of the inequality. 10 + -5 > 0 -10 + 5 < 0 When you divided your inequality you *assumed* that what you were dividing by was positive because you retained the '>'. For any given set of values for which the rest of you analysis is valid, whether this *assumption* is true or not depends entirely on which way round you label the particles. So whether (u1+v1) > (u2+v2) or (u1 + v1) < (u2 + v2) is also entirely *dependant* on your choice of labelling contrary to your conclusion that: "This [ (u1+v1) > (u2+v2) ] is a preposterous result, since even the decision to label them this way round in the first instance was totally arbitrary". That sentence was the entirity of your demonstration that momentum cannot be conserved in an inelastic collision. Regards, Mike. --------------------------------- ALL-NEW Yahoo! Messenger - all new features - even more fun!