1541 testing , heat sink vs, no heat sink

I conducted 1541 heat testing today, heatsink vs no heatsink. No heatsink measured at the middle of each IC, and with heatsink the avg of both sides where the heatsink meets the IC, the difference is staggering. Naturally there will be people who will say that somehow this was measured wrong, thats fine. Ive found that each chip between the middle point and 25% offset to be no more than 1 degree off. Measured off the same points over the same intervals, its clear the heat must be going somewhere. If I were wrong, then the heatsink would be asking like an insulator, increasing the temperature of the chip

With heatsinks. IC#5 would probably benefit if the 2 heatsinks were offset about 4mm to the left from what is shown in the photo