South is Downhill!

Inveterate empiricist that I is, I would make no such claim without proof. After paying attention during three round trips from Olympia to Anacortes, Washington (about 140 miles each way), in a vehicle that tracked the average MPG, the data are clear. Heading north, the mileage is 42.X, while heading south, the mileage is 48.X. (Well, not exactly. The 42.X is the OLY-->ANA avgerage, and 48.X is actually cumulative mileage fro the OLY-->ANA-->OLY round-trip, meaning that the North-south mileage must be more like 54.X MPG. I'm just understating the effect to avoid accusations of sensationalism that a 25% claim would draw.)

You may argue causality, and I must admit that the old "north is up, south is down, and down takes less energy" hypothesis gleaned from my first exposure to cartesian maps seems ridiculously simple-minded. Correlation is not causality.

Further data for the consideration of skeptics:
North trips occurred during somewhat lighter traffic (Alt. Hypothesis: Heavy traffic increases vehicle efficiency?!)
South trips involved the greater urgency of homeward than workward travel (Alt hypothesis" Stepping on the gas increase vehicle efficiency?!)
Prius lies (Alt Hypothesis: Green-inspired data are self-defeating, much like Democrats)

Futher contextual info:
At least two of the three trips occurred during windless conditions.
Olympia point more than 200 feet above mean seal level.
Anacortes point less than 15 feet above sea level.
Driver sensed more downhill headed south than north, but that ain't possible is it?