Hi Georg,
I suppose i should have done this long ago, but I went and did some WWW surfing, since you seem sceptical of my own observations (not to mention arguments). There is no shortage of sites that deal with topics, usually at a technically illiterate level, but there also are some professional-level items as well. Here is one that you will enjoy, I am sure. It is an extract from an abstract:
"It is generally believed that oil samples heat faster in a microwave
oven than do water samples of the same mass. For sufficiently large and
thich samples this conventional wisdom is indeed correct, but this trend
can be far from true in smaller samples. In a commercially-made home
microwave oven, we observed that with decreasing sample size the heating
rate of a water sample increases much faster than that of an oil
sample. At 50 g the heating rate of a water sample is several times
greater than that of an oil sample. Additionally, in studies of
cylindrical samples in a customized oven having a unidirectional
microwave source, the heating rate of water samples smaller than 2.4 cm
in radius is greater than that of oil samples and is a strongly
oscillatory increasing function of decreasing sample radius. Combining
Maxwell's theory of microwave penetration and the heat conduction
equation, we show that this previously unreported oscillatory heating
behavior results from the added power absorbed by samples due to
resonant absorption of microwaves...."
The site is: http://onlinelibrary.wiley.com/doi/10.1002/aic.690400902/abstract
If you have access to the original you might find it entertaining to follow it up.
Other sites of interest in this connection:
http://www.physicsforums.com/showthread.php?t=316981
http://amasci.com/weird/microexp.html
Have fun!
Jon