Update on the previous post: yes, putting a small piece of wet (white) paper towel in the petri dish does increase humidity and growth rate. In not even 24 hours, one of the four plates I had prepared made a nice connection between the electrodes. The other three slime mold blobs on other plates are either wandering around (bottom center photo) or getting nicely settled on their blobs (bottom left and right photos). That’s a nice speed change from the five days I had to wait for last week!
The previous post introduced the idea of “desiccation shock” through Martin Grube’s contributed chapter in Advances in Physarum Machines (2016). I made slightly modified plates, with small bands of aluminum conductive tape protruding from the sides of the petri dishes to easily get a hold of them with alligator clips. Here’s a quick run through of what I did, strongly inspired from Andrew Adamatzky’s work on “Physarum wires: Self-growing self-repairing smart wires made from slime mould“:
First I start with my petri dish plates. I measure a section of aluminum conductive tape long enough to get from the center, then comfortably to the edges of the plates, with some extra (the plates I use at the moment are 90 x 15mm, and I cut aluminum strands of about 80mm in length. By the way, I’m not sure how conductive tape width affects conductivity. I’ve read about 8mm wide strips used here, but mine are 6mm wide, yet they seem to conduct current well enough). I use the extra length to fold the tape on itself, so the plate can stay closed when the alligator clips are connected.
The petri dish can be placed on a ruler, just to make sure there’s a close enough standard space of 1cm between the electrodes. Once that is done —or before— I weight 2% of agar flakes or powder and combine in whatever quantity of water needed (say for 100ml of water, I use 2 grams of agar, as indicated in this helpful post). I use a cheap kitchen scale to do this, and bought my agar agar in a health food store. I put that in the microwave on high for about a minute and a half — but keep a careful eye on it: it’ll start foaming and overflowing quickly. I re-heat until the powder or flakes are dissolved.
Once the agar gel is cool enough to handle —after about 5 to 10 minutes of cooling, approximately — I used a small spoon to carefully drop a dollop of it on each tips of the electrodes. I’ve found pouring the gel in a kitchen squeeze tube to be far more practical recently. Once cooled, the unused gel in it can be re-heated in a matter of 20 seconds in the microwave (still on high) for further use.
Once that is done, I put a small oat flake well inoculated with slime mold on one of the agar blobs:
I also add a new oat flake on the other blob. I spray water on the petri dish lid —to keep humidity high enough—, then fold the dish in plastic wrap and secure the wrap with rubber bands.
I fold again in pieces of black garbage bags to make sure the slime mold stays in as dark an environment as possible. I find folding it in cut pieces of garbage bags also makes for easier transport: they fit really neatly in my camera bag.
Keeping the lid on during experimentation has allowed to light up a LED brightly with lower voltage (23.7 volts in this case).
I can see the LED starting to light at about 12 volts —with current set at 0.02 amps. With this setup, an electronics power supply (instead of a laboratory adjustable power supply) could be used.
Things I’m still experimenting with:
- Whether leaving the slime mould to colonize a first blob for about 12 hours — and then adding the new oat flake on the other agar blob — will make the slime mold eager to get to the new oat flake more quickly and directly (
I’ll publish a post on the amazing escape artistry slime mold is capable of soon: click here for the post).
- How to optimize humidity levels in the petri dishes: putting a small piece of wet white paper towel or sponge in the plates for instance.
- Whether cheap plastic wrap is enough to keep humidity in, or whether parafilm would be better. Plastic wrap is easy to pull off and back on. Parafilm a little less so, but should do a much better job of sealing the humidity in.
I’ve been incredibly lucky to be invited to work on my dissertation and current projects at Matt Ratto’s Critical Making Lab. Since last week, I started bringing the slime mold growing on plates nested in a camera bag, underneath a box of electronic components — musing for a few seconds, every now and then along the way, on the reactions such a find would involve if I ever lost that bag in the subway. I started experimenting on the slime mold with much better equipment and amazing help.
The previous post introduced the idea of resistance in the slime mold. Yesterday I got to examine just how high that resistance could be if the slime mold plates were left open even for a few minutes.
The first time I observed that connection made between the two agar blobs was last Saturday. I inoculated the slime mold on the left agar blob four days before. It went on initially exploring its environment away from the blob and oat flake on the right —as the white traces on the left blob show— before perhaps voting to make a meal of it after all.
I was eager to place that newly grown organic wire into a simple circuit, and see if I could perhaps light a LED with it. I hooked the slime mold wire to a power supply, placed a green LED on a breadboard, and started cranking up the voltage knob. 2 volts: nothing. 4 volts: nothing. 10 volts: still nothing. I kept increasing the voltage until —at about 28 volts— I noticed a faint glimmer coming out of the LED:
28.7 volts! That’s quite an amount of resistance!
Something was wrong. I had read — and been warned— that the resistance of slime mold was somewhere around 3 MΩ. I thought perhaps I had left the petri dish open for too long. But how long was too long? I checked in the Advances in Physarum Machines book, and there it was, at page 27, in a chapter contributed by Martin Grube:
In closed Petri dishes, the air humidity ranges between 95–98 % (as measured by humidity sensors), which appears optimal for growth. As soon as the lid of a containment such as a Petri dish is lifted, air humidity immediately drops to values below 70%, which exposes Physarum to an immediate desiccation shock, unless the lid is closed immediately again to restore tolerable humidity as soon as possible, in the range of several minutes.
So the next step will be to find a way to use the wires while either leaving the petri dish open for as little time as possible, or closed altogether.
I found Andrew Adamatzky’s research very helpful in getting to know slime mold at the intersection of biology and electronics. I’ve been trying to replicate ways to coax slime mold growth to make conductive wires as well. In an chapter contributed to Advances in Physarum Machines (2016) a book he also edited, Adamatzky reported that:
In 25 experiments we measured resistance of protoplasmic tubes on agar blobs. In calculations we assumed length of a tube is 1cm, and diameter is 0.03cm. We found that minimum resistance recorded is 80Ωcm, maximum resistance is 2560Ωcm, median 625Ωcm, and average 825Ωcm. Resistivity of Physarum substantially varies from one experiment to another: standard deviation calculated is 776, which is just slightly below average of Physarum wire resistivity. Average resistivity of Physarum protoplasmic tubes is of the same rank as resistivity of a cardiac muscle of a dog, and skeletal muscles of a dog and a human (p. 234)
That’s a pretty high resistivity! It’s also fascinating that the resistance of slime mold tubes approximates that of our own “skeletal muscles”, as well as that of dogs.
As I started tracking the way slime mold moves along and feeds itself, I’ve been surprised at how it seems to be resisting my own attempts at making it grow into a wire. When I told them about my failed experiments last Tuesday, friends at DIYbio Toronto designed and laser-cut a microfluidic chamber on acrylic. We thought pouring agar mixed with oat powder would help the Physarum hang out in the chamber longer than with non-nutrient medium. And to get it to hang out in the etched pattern even longer, I also put an inoculated slime mold oat at one end of the chamber, and another moist oat at the other end, thinking once the slime mold would have populated that as well, we’d get a happy camper.
So I stayed up late last Tuesday making both non-nutrient and nutrient —oat powder-laced— agar gels and experimented with different setups, the one above among them.
When I checked the slime mold macrofluidic chamber the following day, I got elated: the slime mold had grown one of its filaments and made a connection!
It was a small connection, but I thought with a day or two more, it would get much bigger.
Two days later, as I checked the chamber, the mold surprised me again. This time, it seemed it had retracted the small tube it had grown to reach the new oat flake. Instead of moving into that new one, it made itself more comfortable in its older home-oat-flake:
Seeing that the old flake was all slimy and was perhaps going to start developing unfriendly bacterial cultures, I replaced that flake with another one more than twelve hours later:
On Saturday: still nothing growing in that direction. I thought: “okay, maybe it doesn’t like my new oat flake for some reason”. That new oat flake came from an ACME brand bag of oats: were they processed in some way that hindered the bacterial culture growth the slime mold feeds on? Really not sure. Especially considering that original flake the mold seemed happy to colonize came from the same bag. I switched brands and tried with a new one regardless, thinking maybe the slime mold would be tempted to find its way back to the newer oat flake and make a connection.
But again, I was wrong. Or maybe I was right: perhaps the slime mold did make its way to the new flakes, quickly ate what it found, and decided to go on foraging for better bacterial pasture from its headquarter flake again. Another —completely different—possibility is that any new flake placed in a zone where the slime mold went (and left its smelly waste products) will quickly get slimy as well. The slime mold might not think it fair game to get back to where it once was, especially if its waste products are somewhat toxic to it. After all, the two pictures above and the one below were taken two days apart: this can be long in slime mold land. A better way to visualize this will involve a non-invasive way to take time-lapse pictures of the mold.
Then the final blow came: the slime mold was obviously trying to jump ship when I took this picture of it yesterday:
The slime mold might have fed on —and seemed fed up with— both oat flakes. In retrospect, it looks like such a small setup can feel constraining for the slime mold. In the hopes of getting the slime to stay and bloom in growth in the etched pattern, I instead found that experimental setups with microfluidic chambers better involve short time spans. The attempt at microfluidic domestication resulted in appreciating how born to be wild that critter is!
Since January 2017 I became the happy parent of slime mold. Well, “happy” is how I felt when I took the petri dish that contained — it or them? more on that later — out of the box. Since then, there’s been a lot of mixed feelings: boisterous joys, puzzlement, misunderstandings, flashes of connections to media and communication theories, genuine epiphanies, stress, distress, dejection…
I realized the list of emotions and feelings would lengthen as I explored my ambivalent attempts at domesticating slime mold. The following disordered series of posts all try to account for this. They’re meant to accentuate issues that come up through their cultivation. And they’re also meant to highlight how that cultivation transforms and surprises — never disappoints — ways of thinking and being-with as humans.