Given the texture of the cornstarch mix I think the most important part of this is not only the steady tone but also him breaking up the initial form of the paste at the beginning of the video. For example you can roll it into a ball in your hand, and as long as you keep rolling it its solid state remains. As soon as the motion stops it turns back into a liquid. So with the constant tone it will take on the form of the subwoofer in a solid state until it is broken up by the stick and forced to try to find a new form… Just a thought… Otherwise it may have just been too thick.

If i understand the properties of the past right, made in a big enough batch you can actually run across the surface. But if you stop you sink, so that being said the past is going to act like a solid and just transfer the sound or vibrations straight through it. you might, like the Reverand mentioned, just not make it so thick next time.

You obviously understand the properties of spam, as well...

There was an episode of Mythbusters where they walked on a water/corn starch mixture.

The good Reverend is likely correct, you need to break it up. The frequency probably isn't important, walking would be very low frequency.

The substance in the video seems to be more the consistency of silly putty rather than a thick paste. The mixture you made and photographed looks more the consistency of a thick paste, similar to toothpaste but not quite as thick.

I have done this for open day (I am a science teacher) and I confirm it works. You need to get the mixture to a point where you can pick a bit out of it and roll it into a ball with your hands, and then put it on a table and watch it turn into liquid again. You should be able to stir it slowly, but not quickly.

Then you need a speaker. Any should work really, bearing in mind the next bit: You need to produce a constant low-frequency sound. We had something at school that did the job, but I don't know how you'd produce it at home. Maybe connect the speaker to a PC and use a sound editing program...

Anyway, I put cling-film over the speaker and poured the cornstarch directly into it and it worked fine. When you find the right frequency (you need to be able to vary the frequency) the fingers will climb out of the goo on their own. It is quite a sight. You can pick one off and it'll melt in your fingers. Once they get going you can change the frequency to change the behaviour of the fingers.

Ok for you folks i will explain this.
The cornstarch when mixed with "SMALL" amounts of water will create a STF, or shear thickening fluid. This is a special type of fluid that when stressed will actually harden. This is because when the molecules are stressed in this fluid like state they realign themselves to a rigid and sometimes crystaline structure depending on the materials added. This is actually being used in some experimental Kevlar armor to reduce the amount of damage per layer of kevlar.
Think of sillyputty, you can mold it but if you yank hard it actually breaks instead of flowing. or go throw it on a hard cold surface, it can explode.
Now to get this to work right you need to have it so you can stick your finger in slowly but if you tap the fluid, your finger will not enter. This should be about the right consistency to work.

As a "non-newtonian" liquid, the corn starch liquid acts solid when impacted directly. A sound wave in the air has not enough energy to affect change, which is why in the video the goop is placed on the speaker cone itself.

One optical illusion of the video that is hard to identify is the motion of the speaker. The cone is moving fully three fourths of an inch. "no it's not", you say. The sound wave is at 60 hertz which is very near the frame rate of 59.94 fields per second of NTSC video. This means that the cone movement is being captured in nearly the same position in every frame. You can only perceive a shift in the position of the speaker over a couple seconds. It is most noticeable at the end when the speaker stops moving and returns to rest. That violent interaction is why it works.

I'm attempting to do a similar experiment. Does anyone know if placing a thin layer of glass over the speakers will inhibit the non-newtonian liquid from morphing?

I'm actually teaching a class on how to do this next week. I think you guys have a good start, but I just wanted to throw in a couple details and my two cents in hopes it will help someone.

I use a 2:1 cornstarch to water ratio. Its important to make sure the mixture is actually mixed before putting it in the speaker. The speaker will keep it from mixing if its not completely mixed. Also, I use a program called Tone Generator by NCH Software. It is available free at http://www.nch.com.au/tonegen/index.html. I am using a 10" sub-woofer and driving it with a 350 watt amp. I've had the best results with a frequency between 15 Hz - 30 Hz. When I do it, the speaker moves A LOT which I think is critical to get the fingers to form. I have gotten close to blowing the speaker from driving it too hard at low frequencies, but I still put cling-wrap over the cone trying to extend the life of the speaker as much as possible. Good luck with this.