Vibrational relaxation has been seen in shock waves in propane, isobutene, isobutane, neopentane, and toluene dilute in krypton with the laser-schlieren technique. These experiments cover about 600-2200 K and post-shock pressures from 5 to 29 Torr. The process cannot be resolved in any for T<600 K, or in any for large molecule fraction. All the ultrasonic measurements of relaxation in these at room temperature show characteristic times in the 1-5 ns atm range, corresponding to fewer than five collisions, whereas the relaxation times in the shock waves range from 20 to 200 ns atm, with a clearly defined negative or "inverted" temperature dependence. It would seem the observed slowdown of relaxation with increasing T is simply a consequence of the much increased energy transfer required at high temperature in such large polyatomics when this is combined with a collision efficiency, here interpreted as down, already so large it cannot much increase. The simple method for the extraction of a down from relaxation data offered here by consideration of the energy relaxation equation for Evib=0 appears to be original and should prove quite useful in connecting thermal relaxation data to values obtained from spectroscopy and master-equation analyses. Here it is found that the derived down extrapolate well to room temperature ultrasonic measurements, showing a slight increase with temperature.