Thanks for answering.
Observing higher harmonics has been my first idea as well, however, shouldn't be the first harmonic, i.e. half the wavelength, be the strongest? The observed small scale variations have a very distinct scale. Maybe I should have given more details before. My tube has a length of 0.61m)hence I get for example a resonance for at ~420Hz, i.e. lambda~0.81m. The small scales have a wavelength of about lambda~1cm, i.e. ~33000Hz. Furthermore I did a Fourier analysis of the sound spectrum *outside* of the tube and none of the higher harmonics were very prominent.
Before I performed the experiment I imagined to see the particles moving back and forth rapidly in the antinodes and being more or less motionless in the nodes. However, the particles are moving in these small scale
ripples *perpendicular* to the tube's axis.
Finally, I want to emphazise that these ripples aren't by far an artefact of my experimental design (which is admitably quite simple!) it can also bee seen on all images I have seen so far. As an example I would like you to have a look here:
http://www.physics.montana.edu/demonstrations/video/3_oscillationandwaves/demos/pics/kundtstube3.JPG
I hope somebody can help me with the explanation of these observations.