There’s a bacterium on a diatom on an amphipod on a . . . you know the rest

December 10, 2013 • 10:02 am

From the Smithsonian website:

Once you’ve picked your jaw from the floor, here’s what you’re looking at: the final stop of this zoom, which spans multiple orders of magnitude, is a little bacterium. That bacterium is resting on a diatom, a class of algae that are known for their silica shells. The diatom is, in turn, sitting on an amphipod, a type of shell-less crustacean.

Reddit’s adamwong246 said it best, “There’s a bacterium on a diatom on an amphipod on a frog on a bump on the log in the hole in the bottom of the sea!”

The animated gif was made by James Tyrwhitt-Drake using a scanning electronic microscope at the University of Victoria’s Advanced Microscopy Facility. Tyrwhitt-Drake runs the blog Infinity Imagined.


73 thoughts on “There’s a bacterium on a diatom on an amphipod on a . . . you know the rest

  1. From 1mm scale bar down to 500nm, so that’s a factor of 2000x zoom. Impressive for light optics, but there should be at least another order of magnitude available at the bottom end of the SEM ; possibly two more orders of magnitude. Imaging the cellular machinery of bacteria is routine work these days.
    But that would probably require different preparation techniques to open up the bacterium, which would probably turn the amphipod into unrecognisable sludge.
    Possibly … if they had washed the amphipod with water containing diatoms which themselves had been dusted with bacteria with flagellae … they’d have something more to zoom in upon.
    Nikon sponsor an annual photomicroscopy competition (I’d be surprised if Olympus didn’t do one too – cheap advertising and free plugs like this one). Matthew will probably like the beetles page.

    1. What you’re looking at is signals from the electron clouds of the atoms on the fauna. They are most likely using a back scatter detector. Also, unless my eyes fail me, the magnification is 70k times. Note that the starting magnification is 35x, not 1x. Which seems reasonable for the size of the various fauna.

      1. Note that the starting magnification is 35x, not 1x.

        I meant that the ratio of the magnifications was a factor of 2000 from start to finish.
        I’m still looking for a device that I can keep in my rig bag to provide “objects for scale” when I need to photograph rock samples through the microscope. Magnification ranging from about 1x to about 30x. Suggestions?
        Every so often the clients suddenly turn round to me and start asking for photos of the rock cuttings, but don’t have any clear plans about what they want, how they want it presented, what their purpose is, etc. and every damned client is different. Gets frustrating.

          1. Most of these are eyepiece micrometers, which require the optical details of the microscope to work out the image scale. Harsh reality is that in our industrial setting, the BSc-if-we’re-lucky technicians doing the bulk of the imaging work at 03:00 with 10 minutes total time per sample WILL forget to record the details. And the details will then get lost somewhere between original recording and the final reports being dug out of the archive 6 years later.
            Plus, of course, cheap zoom stereomicroscopes don’t have any reliable way for the user to know what zoom is being used. Or indeed, if the eyepiece zoom is the same as the zoom for the camera.
            The closest that I’ve come is the slides for counting blood cells. But even they don’t clearly indicate which lines are at 1mm spacing, which at 1/10th mm spacing, which at 1/100th mm spacing. And with a budget of BFZ (sterling, dollars, NOK or FCFA), I’m reluctant to speculate my hard-earned cash on a solution that hasn’t been tried by other people first.

            1. I’m not familiar with the microscopic realm, but a common trick in the macroscopic realm for that problem, when no ruler is available, is to use some other common object with a standard size, with coins a perennial favorite.

              Most modern banknotes have microprinting of one type or another as a security feature; putting the sample on top of such a note might do the trick. The ridges on coins are also standardized.

              And are there any common substances whose physical properties ensure consistency? Are skin cells all the same size? Fracture lines in table salt crystals? It’s obviously not a field I’m familiar with, but I think you’ll be able to tell where I’m trying to go with that.


              1. Yep ; so, for an audience containing Russians, Europeans, Americans and representatives from the Beninois Govt … could you suggest a suitable coin in the size range of 0.5 to 2mm diameter, to match the sand grains that we’re really interested in? Not such a simple problem, is it?
                Many years ago the seed of the carob plant was used as a standard of weight – giving us the “carat” as a unit of weight, now standardised as 0.2 gramme.
                The problem is finding something (or things) in the appropriate size range with the size stamped on it in identifiable script. There are grain size comparators and that sort of thing, but they generally have the size marking too far from the actual size indicators to appear in the same field of view.
                Yes, we’ve tried rules of various types. They look really coarse and hard to decipher under the microscope.
                I’ve been looking for something good for literally decades now. I managed to find scales for hand specimens (5-10cm long in ~1cm divisions) a couple of years ago – they’re made for forensics people, and some include colour correlation panels, performing a biavian lithotomb on that problem.

              2. I actually think I can: the American dime. The grooves on the edges are (with rounding) 0.5 mm.

                Of course, not everybody is going to keep American dimes in their pockets, but they’re very, very easy to come by. You can get a bag of them cheap — literally, ten per dollar, which is not much more than you’d pay for washers the same size. Keep a stash with you, and include a few in the sample report. If they get lost along the way, refer the client to the nearest bank. I can’t imagine a bank that doesn’t keep at least a small inventory of American currency and coinage on hand. If it’s a regular or new client, send them a bag ahead of time with your compliments. Maybe make it a nicely-wrapped “care package” with some vials of samples of different grits.

                The other advantage, of course, is that you can feel the ridges on the dime to get a tactile sense of the size at the same time as you’re looking at it.

                I don’t know what type of field of view you’re working with, but the word, “DIME,” is stamped right on the edge and is about 7mm across. If you were sure to include that much in at least one frame, that would eliminate possible future questions over which coin, exactly, is being used as a reference.

                I also don’t know about the imaging tools you use, but, with Photoshop, it’d be really quick to set the scale of the image from even a partial picture of a dime. You’d have on file a standard image of a dime with the resolution set; then, you’d place the unknown image on top, reduce the opacity so you could see both, and stretch and rotate the unknown image so the two line up. The unknown image is now at the same resolution as the standard, and you can use Photoshop’s built-in rulers and / or ruler tool for measurements.

                It might not be perfect…but the perfect is often the mortal enemy of the good….


              3. I can’t imagine a bank that doesn’t keep at least a small inventory of American currency and coinage on hand.

                Many banks will have, or be able to get on order, American notes. I’ve never seen one that has coinage (then again, I’ve never been into an American bank).
                But when your field of view is 2-3mm across in total, you need a small “object for scale”.
                Clients don’t normally ask for cuttings photography until I’m out on the rig, so I have to lash up something out of what I’ve got to hand. Eyepiece projection photography (very difficult to maintain alignment), extra lamps to try to get sufficient illumination. It’s often very unsatisfactory.
                I did have a client once who asked me to organise equipment for routine photography before I went out the rig. That was about 14:00 on a Wednesday, with check-in at 05:00 on the Thursday. Typical oilfield.

              4. Incidentally…you can, for example, tell the Russians that you used an American coin instead of a Russian one because you wanted something cheap and disposable; they should appreciate that. The Europeans I don’t think would think twice about it. Not sure about the Africans.


              5. The point is that no-one actually “knows” the size of any coinage other than the one that they use day-to-day. I’m swapping between UK coins, NOKs and Euros almost on a daily basis, and I still have to read *every* non-UK coin. I don’t have a feel for which coin is which without reading the values.
                Coins do change regularly too. When the US dime is changed to a different size, or abolished, how are people going to find out what the appropriate sizes are. Nope, while coinage is a quick and easy solution for the field (if you’ve got any coinage with you – why carry the weight when you’re out in the hills for a week?), it’s not a professional solution.
                I collect examples of coinage from every country that I visit. Many African states hardly use coinage – just paper money. The coins they do have can be in shockingly bad condition. For the Beninois FCFA, I actually deliberately collected good and bad examples of the coins I saw. The bad ones were polished almost flat.

              6. Actually, I don’t think very many people “know” the size of a coin’s reeding, even if they’re working in retail sales — and it’s the reeding that’s the same scale as the grains. That’s why I suggested the dime and not the quarter; the dime’s reeding is 0.477 mm, while the quarter’s is 0.640 mm — and you were looking for 0.5 mm at the small range.

                And since nobody’s going to have an intuitive feel for the size of their own coinage’s reeding, they’re going to have to have a sample in hand. At that point, the only question is how hard it is to get your hands on a sample. You can get a lifetime supply’s of dimes for your purposes for the cost of a Big Mac, and they’re available pretty much anywhere if the client loses the one you send them. If not at a bank, at an embassy or consulate or military base…or international airport or international chain hotel that’s popular with American tourists or US-registered ship at dock or maybe even from that arrogant overweight guy in the loud clothes with the expensive camera walking down the street right now.

                That is, the advantage of the dime isn’t that it’s known; it isn’t, not at that scale, not even by Americans.

                The advantage is that it’s standardized, manufactured to high tolerances, hasn’t changed dimensions since 1837, is cheap, and it’s not too hard to find a replacement should you lose your last one.



              7. and they’re available pretty much anywhere

                For values of “anywhere” that intersect with “America” ; I’ve been to America and I don’t see any need to return.
                Technically I’ve got the skills to make something to my own specifications using photograph and screen-printing with etching into metal. Or possibly by writing PostScript code directly and then laminating stuff. Printing at 600dpi (a common physical resolution) would allow just under 24 pixels per millimetre … which isn’t going to let me get down to much below 1/8mm. Hmmm, that’s useful ; not ideal, but useful.

              8. 600 ppi is a common laser printer resolution, but inkjets surpassed that a long time ago…but it might not only be a problem to tell an inkjet to print at its true native resolution, it might also be difficult to get it to do so on a transparent substrate without ink spread. Still, you should be able to do better than 600 ppi with an inkjet.

                I don’t know if anybody still has any Linotronic imagesetters in operation, but they used a laser to expose film, and later models did so at 2400 ppi. And they’re all Postscript devices, best I know. Not all that long ago, they were the gold standard in digital imaging.

                Another possible option is their replacement: direct-to-plate printers. Most go to plastic plates, I think, but I’m pretty sure some go to copper. They’re expensive magic boxes that make printing plates from digital files, and any modern industrial-scale print shop is going to be using them.

                What you’re describing, essentially, is making a diffraction grating. Metal ones are going to be expensive, and rolling your own is going to be even more expensive. But commercial plastic ones are cheap; Amazon lists both 500 lines / mm and 1000 lines / mm in various sizes, with a 12″ x 6″ sheet generally costing about $10. Then again, that might be too high a resolution for what you’e looking for.

                But I do still like the idea of dimes…get an hundred of them for ten bucks, and that should last you a long time. If the clients lose them, they can replace them quicker than you can send one of your custom rulers — or, worst case, you just send them another dime.



              9. dimes…get an hundred of them for ten bucks,

                And there I was thinking that a dime was 5 cents. Definite case of incorrectly assuming shared knowledge.
                [SELF : checks coin collection. Ah yes, by deduction from the “quarter” and the 5 cent piece, the one labelled “DIME” is probably either 2c, 10c or 50c. But you have to work it out. It’s not a currency that I’m familiar with, to repeat an earlier point.]
                Inkjet printouts have a distressing habit of bleeding when they get wet. busy (mud) logging units are wet places. We put up with it for years until most clients started supplying internet connections to where we were rigging up, and then we just stopped sending printers at all. What couldn’t be done by email we’d do with pencil (NB : not pen!) and paper, or create a PDF and drop it onto a printer when in one of the company offices.
                Diffraction gratings … hmmm. At that sort of price, that’s worth thinking about. But I’ve got other stuff to do today as well.

              10. Damn. Just when I thought I had a handle on “standard” measurements, I learn there’s even more to it. And you’re either not kidding or you’re quite fast with the Wikipedia revisions.

                I also didn’t realize that barley grains were so big. The wheat berries I’m familiar with are…here, let me grab some from the freezer…yes, they’re all between 3/16″ and 1/4″.


      1. So they work by magic?
        Little elves would count as cellular activity.
        (Oh, and BTW, “bacteria is a plural ; the singular is “bacterium”.)
        For a list of components, there’s the “pili” (hairs) outside the cell wall and flagellae in various combinations and patterns ; the cell wall, surrounding the plasma membrane. Then internally there is some structure – the nuclear material is often noticeably concentrated ; there are ribosomes – where RNA is translated into proteins ; depending on your bacterium’s type and life stage there may be food or gas storage vacuoles. Occasionally magnetosomes (iron oxide granules that respond to magnetic fields).
        There’s a significant amount of structure inside bacteria. But you need to break open the cell wall – which is often not easy, and the process would likely destroy the diatom and the amphipod, ruining the picture.

        1. Thanks for correcting “Eric”s fatuous notion that bacteria have no biological activity. Otherwise, we’d be 20 feet deep in undecomposed organic matter.

  2. “So nat’ralists observe, a flea
    Hath smaller fleas that on him prey;
    And these have smaller fleas to bite ’em.
    And so proceeds Ad infinitum.”

      1. “The bigger fleas, have smaller fleas,
        Upon their backs to bite em.
        The smaller fleas have smaller fleas,
        And so, Ad Infinitum

        (That’s how it was taught to me as a kid)

      2. “The bigger fleas, have smaller fleas,
        Upon their backs to bite ’em.
        The smaller fleas, have smaller fleas,
        And so, Ad Infinitum.

        (That’s the way it was taught to me as a kid, but I like the other version, too.)

          1. from Wikipedia:
            “The Siphonaptera” is a nursery rhyme, sometimes referred to as Fleas.

            Big fleas have little fleas,
            Upon their backs to bite ’em,
            And little fleas have lesser fleas,
            and so, ad infinitum.

            Sometimes a second verse appears, with lines such as

            And the great fleas, themselves, in turn
            Have greater fleas to go on;
            While these again have greater still,
            And greater still, and so on.

            The rhyme is closely based on lines by Jonathan Swift from his long satirical poem “On Poetry: a Rhapsody” (1733):[1]

            The vermin only teaze and pinch
            Their foes superior by an inch.
            So, naturalists observe, a flea
            Has smaller fleas that on him prey;
            And these have smaller still to bite ’em,
            And so proceed ad infinitum.

            Lewis F. Richardson adapted the poem to meteorology, specifically discussing fractal wind patterns:

            Big whorls have little whorls
            That feed on their velocity;
            And little whorls have lesser whorls
            And so on to viscosity.

  3. Now, all we need is to keep zooming in until we see the virus on the bacterium on the diatom….

    One of the tidbits that keeps blowing my mind every time I remember it or that I stumble across it is that we’re smack dab in the middle of the size of the Universe. That is, we (roughly one-meter scale) are about as much bigger than Planck Length (roughly 10e-35 scale) as the observable universe (roughly 10e26 meters) is bigger than us.

    To put that bacterium in perspective, you are about as much bigger than it as the entire state of California or the whole country of Italy is bigger than you.



    1. Our place in the “zone of middle dimensions” (can’t remember who first called it that), blows my mind too. For some reason, I’ve never liked using a microscope; it’s just too weird that there’s so much there there.

        1. To me the micro is just plain fascinating and I could go all the way to the Planck length without any problem…on the other hand thinking about all the huge stuff going on out there, neutron stars, background radiation, black holes, galaxies colliding and so on makes me just feel so worthless that even just a glimpse of the night sky is enough to faint.

    2. No, we aren’t anywhere near the middle of the size of the universe. The ratio of our size to Planck Length is 10e9 times that of our size to the observable universe. Which means that we are a billion times closer to the big end than to the little end!

      1. A billion might seem like a big number, but it’s not in this context.

        Here it is graphically. Planck scale at the left, the observable universe at the right, us in the middle:


        Most people are going to have to look closely to be able to tell that we’re a bit closer to the right than to the left.



        1. Ah, perhaps I misunderstood when you originally said “smack dab in the middle”. We’re just not that special.

          1. If I gave the impression that I was trying to suggest that we’re at the center of the Universe, especially in an anthropocentric sense, please forgive me; that’s not at all what I was going for.

            We are, of, course, astoundingly special. As Richard Dawkins put it, we are going to die, and that makes us the lucky ones. But of, course, our specialness is one of accident, of the pothole perfectly fitting the puddle, of the Lottery Winners Association Book Club having an extraordinarily high net worth.

            The macro scale is the scale where chemistry and Newtonian mechanics overlap, and that would also appear to be the only domain where things get “interesting” enough in our universe to support the type of complexity necessary for ambulatory computational devices smart enough to ponder these sorts of questions. Whether there’s any significance to the fact that said scale is right in the middle…well, I don’t know that that’s a question that anybody’s put any serious thought into.

            Of course, lots of things in statistics take the shape of Bell curves, which concentrates values in the middles of ranges. If multiverse theory is correct, that’s where I’d put my bet as an explanation.



    3. Hooray! I have always wondered about the “middle”. I speculated it might be the atom. Or that the middle is the place where the macro world and micro (quantum) worlds meet. (where is that?) So the middle is us? Man? or as you said one meter which is close enough for me. Thanks for your post.

      1. No, as I pointed out, we’re not in the middle. The range is from 10e-35 metres to 10e+26 metres, so the middle point is around about 1/50000th of a metre, or 20 microns.

  4. Scanning EM is so cool. Back in grad school I would zoom in onto freeze-dried crayfish legs (my research subject), and would often find diatoms, stalked protists, etc. A person can get mighty distracted by such such things.

    1. Scanning EM does not directly produce a color image, since the image is built from electrons scattered from the surface of the specimen. They can be secondarily colorized, but the colors are not real.

      1. That…and, have a look at the scale on the last frame of the image.

        It reads, “500 nm.” I’m pretty sure that applies to the length of the whole scale, which would mean each tick mark is 50 nm.

        Green light has a wavelength of 500 nm. Human vision extends from about 380 nm (near UV) to about 750 nm (near IR) — and the extremes are pretty dim…in practice, it’s closer to say, about 425 to 650 for the limits of what you’re really going to see as opposed to what you can make out in the right conditions.

        That means that you’re seeing details on that bacterium that are much, much, much smaller than the wavelengths of the light that your eyes can see. At that scale, color ceases to have any meaning.



    1. who says they’re trying to prove anything? why can’t you just enjoy the beauty of seeing something in a way you’ve more likely than not, never seen before?

  5. I’m a blue collard, red neck kinda guy, working hard for my family, and even I can see the beauty of this. Dont know what these creatures are, but I can only imagine what may be riding on this bacteria’s backside!

Leave a Reply