Scientist Stories: Eric Betzig, Historical Connections Between Astronomy and Microscopy

Good evening.

I'm Jerry Rubin. I'm the Executive Director of the Genealey Farm Research Campus, and I'd like to welcome you to tonight's presentation in our ongoing series, Dialogues of Discovery. Now when you walked in, hopefully you all got this little booklet here that describes Eric Betzig's background, and I'm not going to go through all that, But if you've read it and you've been to some of these lectures before, you probably saw that his background's a little more unusual than the typical academic scientist who goes from one university to another. You might have also noted that it says that he started here as a group leader in 2005. And if you lived in Loudoun County for a while, you might know that in 2005, the only people working in this building had hot hats and the building didn't open until the fall of 2006. So what really is the story here? Well, it turns out that in 2005, while we were under construction, I got an email from a man named Eric Betzig who— sorry, a man named Eric Betzig who I have good— it's my wife telling. Anyway, good coaching. An email from someone, Eric Betzig, who I never heard of, saying, you know, I've been out of science a while, but I used to work at Bell Labs, and I'd like to get back into academic science, and I heard about Genelia Pharma from a friend, and how do I apply for a job there? So I knew some people who had worked at Bell Labs, some of the people who were advising me, and I sent this email to him and said, you ever hear of this guy Eric Betzig? And he said, Wow, you know, Eric Betzig was a legend in his time at Bell Labs, and then he just vanished. One of them said, I'm glad to get that signal because I thought maybe he was the Unabomber. So, and now a legend is not normally a term that scientists use to describe one another, but Eric really was something special. So he didn't drop out of science, as it says here. Because he wasn't any good at it. In fact, in 1993, which is a couple of years before he decided to leave academic science, he was a recipient of an award that's given annually by the National Academy of Sciences for, quote, to recognize innovative young scientists and to encourage research likely to lead to its capabilities for human benefit. The award is given to a citizen of the United States, preferably no older than 35 years of age. So this is like someone, if you imagine, use a baseball analogy, you had just won the Cy Young Award and you announced to the world that you really didn't like playing baseball anyway and you were going to go do something else. And no one knew of his scientific colleagues where he was or what he was doing. So we invited Eric to come for an interview and I had no trouble convincing these curious individuals to come. Participate in the interview. And to make a long story short, 6 weeks after we did this interview around the office buildings that are just up there, we had hired Eric. We were supporting his laboratory, which happened to be in an industrial park in somewhere in Michigan, because we had no lab here and he lived in Michigan. And so we just rented him space, and he worked there for a year until we opened. So he's really one of the first people we recruited here as a scientist, the first person who was a paid scientist on our payroll, so it's a particular pleasure that he's giving this talk. You also might notice that he did preliminary work that led to his job in the living room of his collaborator and good friend Harold Hess, who also now works in this building. And you might thought, I thought budding young inventors worked in garages, not living rooms. So I want to convince you they really did work in a living room, and I can show you photographic evidence of this. This is the Palm prototype. Now you can tell this is not a Janelia farm by the style of the furniture, and you could probably also tell that Harold wasn't married because my wife would never let me do this in my living room. Anyway, so it's a pleasure to welcome up to the podium who's going to tell us about magic, all things great and small, historic connections between astronomy and microscopy. Eric.

Well, thank you, Jerry, for that introduction, and thank you for rescuing me from obscurity in the end. It's been a fun ride anyway so far. So, and thank you all for coming here today. I've always been very interested just as an observer from afar from my in astronomy because when I was an undergraduate I was always thinking of taking that path instead of the microscopy path I took. And many times since then I've kind of went, oh my God, why didn't I become an astronomer? But that's just because the astronomers are much smarter than the microscopists and we always end up stealing from them. And that will be a common theme throughout this talk. Ever since the beginning of recorded history, men have been asking themselves what is their place in the universe. So they'd look up and they'd see the stars in the sky and they'd say, how far away are those things and what are they? Or they'd look down at their feet and they'd say, is a grain of sand the smallest thing there is or is there something smaller? After many, many thousands of years, today we have a very good picture of exactly where we are in the scheme of things. And it's very humbling. It's so humbling that it's impossible for you to really intellectually understand exactly how big the universe is, but I'm going to try. So this, for example, we're normally about 2 meters tall. Here's a meter big patch of leaves on the top of a tree. How big are we compared to the molecules inside of our body? If we start zooming in by factors of 10, 10, first you see a leaf, then the veins in the leaf, then little clumps of cells. Zoomed in a little bit more, there are the individual cells. Now we're seeing a nucleus inside of a cell. This is now a million times magnified versus the scale of our bodies. Now you start to see the super organization of DNA in the nucleus, the DNA itself, and the nucleotides that compose the DNA, which are the individual molecules. So in other words, we are about a billion times bigger than a single molecule. But where are we on the other side? So again, if we go back to the scale of us and start zooming out, first you start to see a bunch of trees, then you see some buildings, then you see more buildings, then you see Tallahassee, Florida in this case, Southwest Florida, the Southeast US, the Earth. Now we're the Earth farther away. That's the orbit of the moon, a factor of a billion the other way. In other words, if you go a billion times out as opposed to a billion times in, you only get as far as the orbit of the moon. Okay, so as you know, the universe is a heck of a lot bigger than that. How big are we compared to a galaxy then? So going from there, another factor of 10, that blue thing is the orbit of the Earth. Now you start the orbits of Venus and Mars, the inner planets, the outer planets starting to come in. The whole solar system. This is now the sun at a trillion kilometers. Now we're at a light year. You can still see the sun. Now it's lost in the stars at 10 light years. And we have to zoom out more and more and more to get to the whole galaxy. So by the time we zoom out another factor of 10 to the 13th, we're at the galaxy. But that's still only 1/10,000th of as big as the observable universe. So what I'm saying is that we are 10 trillion or 10 million million times closer to the size of a single molecule than we are to the size of our galaxy. And our galaxy is nothing in the vast realm of space. So it is really hard to really appreciate exactly how insignificant we are. So whenever I start to think that I'm really hot stuff, all I have to do is think about that and realize that I am nothing and we are all nothing. And that's just the way it is. So how much of this can we see with the naked eye? So it turns out, I'm sorry, this is about the only physics you're going to get, but if you think about the eye, it has a pupil which determines a certain diameter of light that comes in. And if I want to see something close, I move up close next to it. So the resolution, or my ability to see two close-packed things, gets better as I get closer or if my iris gets bigger. Likewise, for things that are infinitely far away, or at least to us, like the stars, all that really matters is the size of the iris. And if we had bigger eyes, we could see better. Okay, but just like in a microscope, if we had a bigger lens and could place it closer, we can see better. So in both directions, whether we're looking at the realm of the small or the realm of the large, Making a bigger instrument diameter-wise gives us better resolution, okay? So, but what that means going back is that the resolution we have with just our naked eye is too big by a factor of 50,000 to see single molecules. It's good enough to see about a single hair. But on this big side, it's too coarse to be able to see even Neptune as a disk by about a factor of 50, and we can't even see most of the of the galaxies in space. So how do we know all of the stuff I just told you at the beginning about the scale of things? We know it because we can make instruments that have bigger diameter pupils to allow us to see more. In microscopy, that means lenses of big diameter that are placed very close to the surface, only a few hair widths away from the sample. In astronomy, that means making bigger and bigger and bigger telescopes. So this is sort of state of the art right now, the 10-meter scope at Keck in Hawaii. That's a person sitting in the middle of that telescope to give you a sense of scale. So I'm going to tell you a little bit about the history since these instruments were basically invented at about the same time, 400 years ago, the microscope and telescope, and explain how improvements in that technology led to the understanding we have today of the scale of the universe and our place in it. So optics, it goes way back to pre-Christ times. So an example is Archimedes studied optics. There's a probably false story that during the siege of Syracuse where he lived, that he came up with the idea of using a bunch of mirrors to reflect the sun and burn the ships of the approaching Roman army. Mythbusters looked into that a number of times and basically it was not all that impressive. You can kind of char it, but let's just say it's not going to work too well, as was found out by Archimedes himself because Syracuse fell after 2 years and a Roman soldier killed him. So I guess that death ray didn't work too well. But another technology, another 9 centuries later was the— or actually 11 centuries later was when a Muslim scientist found a way of taking sand and making very pure, clear glass with it, which could then be formed into a hemisphere to make what was called a reading stone. So all the monks who were working in their little cells in the dark or in candlelight reading the Bible, and their eyesight going to hell, could start to read better by using these reading stones and see the Bible. Eventually, that led by the 14th century in Italy to the development of eyeglasses, which was a huge boon for people, particularly the wealthy who could afford these things to be able to see for longer. And then we come to the end of the 16th century when there were almost simultaneously two discoveries. Zacharias Janssen realized that if he took not just one lens to look at something, but took two lenses together, he could see much better than just a single lens alone. And that was the formation of the compound microscope. At the same time, the legend has it that this eyeglass maker, Hans Lippershey, was working one day and two boys took a couple of his spectacle lenses, went outside and they were playing with them and they put them together like this and they could see a weather vane across the harbor much better than before. Or they told him and he went, "Eureka!" and that was the birth of the telescope. So who knows if that's true, but at least that's the legend. So that's— so Lippershey then immediately started to make these telescopes. They became a sensation across Europe and they became a sensation for the European explorers who were going to the New World because they had much better ability to see across distances. One of the guys who first heard about it was Galileo Galilei, and he immediately went like that because he said, oh my God, of course that would work. And so he started to make his own telescopes using the same type of technology, but because Galileo was a complete genius and one of the most talented men in history, he quickly was able to improve the resolving power of the telescope by a factor of 10. Pen over what lipper she had. And as soon as he started pointing that instrument to the heavens, everywhere he looked he made a discovery. He was able to see that the moon was not smooth, but had a rough surface of craters. He was able to see that Venus wasn't just a star, but it was a sphere that had these phases just like the moon. He was able to point it at Jupiter and see that there were these little dots around Jupiter in a line, and as he looked over time, these dots would move., and he realized those were the moons of Jupiter. That got him into a lot of hot water because at that time the Earth-centric system was still in place and believed by the church. And so as he started to promote a heliocentric system like Copernicus, this eventually led him to the Inquisition and house arrest. So that's what he got for his troubles in that case. So this is the first example. The microscope did almost nothing for about 50 years. The telescope, once it was in Galileo's hands and Kepler's hands and many others, started to make discoveries immediately in the 17th century. But finally by the second half of the 17th century, microscopy started to catch up. First with a guy named Robert Hooke. So Hooke created a compound microscope of of his own design with a much better illuminator. And Hooke was also trained as an artist, so he could look at these things for long periods of time and make these beautiful illustrations, which he then eventually compiled into a book called Micrographia. This was the first bestseller of any kind across Europe. It was a real sensation, not just with scientists, but with the lay public as well, because of pictures like this flea. At the same time this book was published, the great plague of London was still ongoing, and these fleas were the carriers of that plague. And people finally realized for the first time through this type of work that science had a real practical benefit to be able to understand what was going on. And so it was really a sensation. Unfortunately, Hooke himself was by all accounts a really horrible person. And everybody hated him. He was a genius, but he made sure that everybody knew about it. And he was paranoid, and he believed that everybody was stealing his ideas left and right. And the guy he pissed off the most was Isaac Newton, who was his contemporary, and probably the only man smarter than him. And Newton did everything in his power to erase all mentions of Hooke from history. There is no pictures of Hooke that you can find, because Newton destroyed them all. So Hooke has never really gotten his due, but as time has gone on and more has found out about what he did, people are beginning to really appreciate his work. Hearing of Hooke's work, a guy named Antonie van Leeuwenhoek in the Netherlands was able to make his own type of microscope by a different principle. Instead of using two lenses, which should have been better, he came up with an idea that was just using a single lens. But I'll tell you in a minute about his idea. But with that microscope, he could see much better than anybody else. He could see rotifers. He could see that veins and arteries were connected by capillaries. He even got to the point where he could see single-cell bacteria. And this was 150 years before anybody else could see the same thing. That was so crazy that people disbelieved him and the Royal Society of London had to send people out to look at it to see if this was true or if he was seeing things. And they were able to confirm it, so now he's called the father of microbiology. The thing though is that he was a secretive bastard and the way he made his lens, he told nobody else because he realized that he was the seeing man in the kingdom of the blind and he'd have competition for his work if he said how he did it, and so he kept it quiet, which means it went away with his death, which means microscopy basically went sort of dormant for another 100 years because he kept this to himself. But his trick was basically to take, instead of polishing a lens, which is what everybody else did and what he told everybody he was doing, instead he would take a glass rod, heat it up, and pull it like a piece of taffy until there was a long hair of glass. He'd put that in the flame and it would curl up into a little ball, a perfect ball, and he'd break that off and that was his lens. And they'd be only like a millimeter big, but they could see so well that they got within a factor of 4 of the fundamental resolution of microscopy that wasn't reached for another 200 years. But again, it lived with him and died with him, but he made amazing discoveries. His discoveries, which he sent in letters to the Royal Society for a number of years. So the telescope was also marching along, but at a point it hit a brick wall, and the brick wall was because the lenses that were used in the telescope don't treat all colors of light the same. And so red light is not bent as much by a lens as blue light is.. So the point of focus of these light rays is not all the same. And so when you look through a non-chromatically corrected telescope like at this weather vane, you can see these different colors and the smearing due to this aberration problem. And it really limited what you could do with these microscopes, or these telescopes. So the way the astronomers dealt with this problem is that sort of displacement between the different colors and how crummy it is can be improved by just using a lens that has a really long focal length so that you don't bend the rays that much to get to the focus. So now you start making lenses of ridiculously long focal lengths. And this was therefore the era of what's called the aerial telescope. They would make telescopes that are hundreds of feet long and hang them in contraptions like this. And put one lens at one end and one lens at the other end and be able to minimize chromatic aberration that way. One guy doing this, Christoph Huygens, made what was an aerial telescope. He got rid of the superstructure altogether, hoisted the one lens up a flagpole, yanked on it while it was on a ball joint, and looked with the other lens. With that, he was able to see the rings of Saturn, be able to trace how the rings change as Saturn goes about its orbit. Made many, many discoveries. The resolution of what he could resolve in the sky was not to be improved upon, although telescopes got much better by metrics, light collection and everything. But in terms of resolution, the resolution he got was not to be improved until about 20 years ago. So that's how good they could do with these kinds of things. But man, they were a pain in the neck to use. You had to have good eyes and steady hands. So the problem of chromatic aberration could be corrected by another way. The guy who really understood chromatic aberration was, again, Newton. 'Cause Newton was the guy who really saw by putting white light in a prism you get this spectrum, which is really the origin of this problem. And he said, well, that's if you're refracting light through glass, but if instead you reflect it off a mirror, all the different colors reflect at the same angles. By making a curved mirror, you can completely eliminate that problem. So Newton created the first of what's called not a refracting telescope, which uses glass, but a reflecting telescope that uses mirrors. And by doing that, he was able to make a telescope that was much smaller that performed pretty darn well, not as good as the best Aero reflectors, but for its size and cost, much better. The problem is, is nobody back then knew how to make really good mirrors, so they used an alloy of tin and copper called speculum. And you would have to polish this over and over again to get the right shape, but then it would immediately start to tarnish. And then you'd have to go and polish it again. It's hours to get the right shape. And he would do it with a spherical surface to get what he was doing. But a spherical surface has another problem, which is another kind of aberration called spherical aberration, is that with a spherical surface, all the light rays that bounce off the mirror don't come to the same point. However, if you can create an elliptical surface, which is much harder to do. If you've ever taken a ball of Play-Doh and rolled it in your hands, you can see you can make a pretty good sphere without a heck of a lot of trouble. But if you had to make something that's kind of a parabola or an ellipse, it's really much harder. But this guy Hadley, by the early 18th century, was able to do that and create a parabolic reflector telescope that was only about this big versus something that's across this whole room or more for an aerial telescope and had roughly equivalent performance. So now all of a sudden the reflector took over versus the refracting telescope. And this was the beginning of the end for these aerial telescopes that dominated then. The next guy to come along was William Herschel. And Herschel was actually trained as a musician. One of the things that keeps coming up in this history is that I may not be the only guy who has a weird history. These guys all have weird histories, okay? Very few of them were classically trained as scientists. Many of them were merchants. Many of them were, like him, a musician or coming from other fields, and astronomy or microscopy was just a hobby. It was a hobby for him that became an obsession. And so Herschel was basically the guy who was willing to spend lots of time by polishing these bigger and bigger mirrors in these elliptical shapes to make better and better telescopes. He would literally work 18 hours a day going around, his mirrors were this big, and so he'd go around in a circle like this with a pole in the middle with his polishing pad going around the mirror. And you cannot lift your hand because then you completely lose the smooth surface you're trying to get for 18 hours while his sister Carolyn would feed him his meals as he went around doing that. When those 18 hours were over, he would go back to one of his earlier telescopes, because he'd do that during the day. And then at night he'd make his observations with the other telescopes. But with those observations, he cataloged more of the universe than any other human in history. He found Uranus. He found moons of Saturn. He found the first binary stars. He cataloged thousands of binary stars, nebulae, the ice caps of Mars. Everywhere he looked, he made a discovery because his telescopes just rocked. Compared to anybody else's. The biggest one he made was this one that was called the 40-foot telescope here. It's this big. This would be him going up in this cage to look in right here at the image that he would get. That was hand-built by himself, okay? So this is the scale of what this guy did. So that's great, but again, these speculum mirrors were a real pain in the neck because they kept tarnishing, which meant Herschel had to spend it would have been another God knows how long to be able to polish his mirrors and the others. So, but there is a way with the lenses to get past that problem of the chromatic aberration. Newton said it was flatly impossible. There was no way to do it. This is one of the few cases where Newton was wrong. And that there was a guy, this Chester Moore Hall, he was actually a lawyer in England, and he realized that if you took two materials of of different bendability of the light rays called the refractive index and cement them together, one positive, one negative, you could create a canceling out of that effect to get all the light to go to the same spot. So he was just a lawyer again, so he didn't build these lenses himself. He contracted that out to two companies to do the two different parts of the lens. They subcontracted it as it turned out, to the same subcontractor. That subcontractor said, "Oh, I know what this guy is doing." And so then that subcontractor told another guy, and that other guy 30 years later patented the idea even though it was this other guy's idea. And so that's Dolan who patented it, and then he got rich off of it. But, you know, Chester Moore Hall, he wasn't that much interested in either the money or the science, so I guess everything kind of worked out. But that's That's basically the start of these achromatically corrected telescopes. However, it was done completely empirically. So in microscopy, the guy who started to figure this out in a quantitative way was Joseph Lister. He was actually a wine merchant, but he was also an amateur astronomer and also a good mathematician, and he was able to figure out the math of exactly how you would have to put things together to make a good microscope objective that would correct for chromatic aberration. He also figured out that by changing the spacing of the lenses ever so slightly, you could correct the problem of spherical aberration that occurs in microscope objectives. And so now he was finally able to get to the point, by all of these tricks, of getting a microscope that was as good as the microscope that Leeuwenhoek made with that little single ball 150 years previously. And then this technology his nephew commercialized with a guy named Ross, and then this started to be the beginning of commercially available microscopes for many research uses. But it still wasn't perfect, and so the microscopes continued to improve throughout the 19th century until this guy came on the scene, Carl Zeiss. Zeiss first started to make these simple lens microscopes like von Leeuwenhoek's, and he started to sell those pretty successfully on the order of 50, but he said that I know I can do better if I make a compound microscope like Lister's. So he started to do that, he was very successful, started to sell hundreds of those, but this guy was a demanding, demanding man. He was a real businessman and he He had a whole factory of people hand-making these microscopes, and when they weren't up to his specifications, he would literally take a hammer and smash it in front of the guy who was assembling it. I mean, he was an angry, angry guy, but he said— but he was angry at himself because he realized that he was just empirically getting this stuff, and it wasn't really a science, and he didn't really know how to make it a science. He lived in Jena, Germany. University of Vienna was nearby. There was a young physicist named Ernst Abbe who had started there a few years earlier, and he convinced Abbe to start looking at the problem of how to find a mathematical basis for how to make a perfect microscope objective. Abbe succeeded in that task spectacularly. He found out exactly what you need to do, and furthermore, he found that there actually is a fundamental limit to the resolution you can get in the microscope beyond which it doesn't matter what you do with this technology, you're not going to go further. This is known as Abbe's Law today, and it kind of makes sense. It says since light is made of waves and waves have a wavelength of whatever, you can imagine the waves like the fingers on your hand, and if they're trying to look over a sample to image that sample, you probably can't see anything much finer than the wavelength of the the light that's interrogating it. And basically he found mathematically that this was the case, and he was able to prove that. But even with all his brain power applied to Zeiss, the microscopes that they made according to Abbe's principles still weren't as good as Zeiss was willing to have and not as good as theoretical limit that Abbe had found. And so Zeiss recruited a chemist who worked on the chemistry of glasses, Otto Schott. And Schott was the guy to make these really pure, really different types of refractive index glasses that opened a whole new world, not just for microscopy and astronomy, but for lighting of street lamps and so forth. These three guys got extremely rich between— particularly with the technologies of glassmaking that Schott created. And with Schott's glass, they were able to consistently produce microscopes that are as good as any research microscope of this technology that you can buy even today. So they really were at the limit. By the end of the 19th century, that's where we were. And again, in the 100 years since then, optical microscopy has become a complete workhorse for discovery in biology because it's non-invasive, so you can look at living things. With this type of technology. It's a big player in what we do at Janelia. It's a big player everywhere. It's a very critical technology everywhere. In the meantime, with the advances in creating these achromatic lenses, astronomy wasn't going to sit by and do nothing about it. Astronomy was going to catch up. The problem was that initially when they came up with these achromatic lenses, the glasses for the second lens was a very difficult type to manufacture and difficult to make in a big size like you would like for a big telescope with a big pupil to see well. But eventually by the 1850s, they figured out how to make very big blanks of all of these glasses, and then by the late 19th century, this was the time when people could make really big lenses like this lens in this telescope here. And basically now the reflectors, which had taken over from the refractors, now the reflectors were dead because nobody wanted to polish those speculum mirrors over and over again. Now they could get just these lenses made once. These instruments, which are 100 years old, are still in use today and are still producing real science, okay? So that's how significant this was. Was. This is also the dawn of when you can start to see we're getting beyond the individual guys working in their basements or laboratories or living rooms if they're not married or what have you. But it's still philanthropy, just like us here from Howard Hughes. Philanthropy was very big in the ability to make these types of instruments. The next major innovation for both fields was the development of photography. Because if you're just looking with an eyepiece, you can only look for so long. Things are changing, you're blanking. You can't sum up all of the information that you can if you just put a photographic plate there and wait for a very dim object to be exposed for a long period of time. And so by the late 19th century, people were able to get these great images of comets, galaxies, and Saturn with these large refracting telescopes not using eyepieces, but now using photographic film instead. And it started to have an impact in microscopy too, to be able to record things rather than having to just describe them or try to make drawings of them. Then finally, in again the 19th century, a process was developed to be able to not have to use that damn speculum metal anymore, but to use a glass blank instead that you could put a silver coating on. And then if you did that, then the silver coating didn't tarnish anywhere near as quick, and if it did tarnish, you could strip the coating off without harming the curvature of the glass underneath, and you could reapply it. So now all of a sudden, the reflectivity reflectors started to take over because those big refractors I showed you, like the Yerkes refractor, that telescope was at the absolute limit. It took 18 tries for the manufacturer to make the lens. And many of them, as soon as they start to try to lift it, gravity would cause it to either sag or break. And so it was right at the very hairy edge of that technology. But now, because you could make a much a much lighter weight mirror blank out of the glass and then put silver on it, you could make something much bigger. And so the reflector started to take over again, and now we're starting to see things that start to look like modern astronomy. The Orion Nebula, Great Spiral Nebula. This is the Hale Telescope that was developed at Caltech. This one you can actually rent. It's up on Mount Wilson near Caltech, and you can rent it for the night. A choice telescope to be able to do that kind of thing, just for the general public, and the Hooker Telescope. But the big milestone in the era of the great reflectors was the 200-inch Hale Telescope dedicated right after World War II. This is the primary mirror of that telescope. This is the dedication. That tells you the scale of things here. This is the mirror here. Where they've stripped off the coating to reapply a new coating. And the reason they could go to these even bigger mirrors is because a guy at Caltech came up with a way of depositing aluminum that would last even much longer than the silver and have a much better, higher reflective coating. And so this instrument was for 50 years almost the world's largest effective telescope and was the workhorse of astronomical discovery. Discovery. When I was in high school, I had these pictures on the wall of my bedroom because to me this was what I wanted to do. This was amazing at that time. Again, this was the guy who ruled the roost for 50 years. The next revolution was beyond photography, was digital imaging. So being able to use things like your cell phone has or CCD chips. In an inspired choice, the Nobel Committee in 2009 gave the Nobel Prize to these two guys from Bell Labs who invented the CCD chip, not for imaging but for actual data storage. And then the imaging applications came later as a spinoff. That's the way basic research is. You never know how one thing is going to spin into another. But basically this has transformed completely astronomy and microscopy. This is a picture of an early star field, early CCD image where he compared the film and its sensitivity. And this is good film compared to what you would see if you imaged it with a CCD. So the fact that you could digitize and then process the data and do all of this was just revolutionary. And it was revolutionary for microscopy as well because now you could take an image that had very low contrast like that, digitize it, play with it, heighten the contrast, and see all sorts of things you couldn't before. You could use extremely low light levels, so now you could look at the dynamics of living cells under the microscope without cooking them by having too much light on them. By the early '80s, this was starting to really transform what we could do in microscopy. So that's pretty much where things stood. The question is, is beyond the Hale Telescope or something like that, are there other ways of seeing things? Or beyond the Abbe limit, are there ways of seeing things on an even smaller scale? And the obvious one is if we look at something that isn't infinitely far away, or to our abilities infinitely far away like the stars, something closer like the moon, the angular resolution goes just inversely with the diameter of the telescope. But the spatial resolution, as you get closer and closer, since this number goes down, the ability to see things closer and closer gets better as you get closer and closer with the telescope. So if I can put my telescope not on Earth but in orbit around the moon, then I can see things much better, right? By after Sputnik, after a few years, we're now starting to put our cameras and our telescopes close to planetary bodies. This is the moon in this case in 1967. This is Saturn, which first was visited by Pioneer 1 in '79. Today the Cassini orbiter is still going around there. These pictures are what Cassini can do today. This picture looks like it's an artist's conception That's a photograph, okay? What you're seeing is Saturn eclipsing the Sun and the Sun's rays behind it illuminating all of the rings. That tiny little speck there is Earth. So this gives you, again, the scale of how insignificant we are. And this is our cosmic neighbor, really close to us on the scale, and this is us. So Cassini is still producing data like this every day, and if you're interested in this kind of thing, or what the Curiosity rover is doing on Mars now, or the Mars Orbiting Observer, or the Messenger probe around Mercury. We have stuff all over the solar system that's daily taking beautiful data. And it's, it's, there's a revolution that is continuing to go on in astronomy and planetary astronomy, which most people completely ignore. But it's really amazing what's happening today. So can we apply this principle of getting even closer to somehow getting past this supposedly fundamental limit that Abbe had to the resolution of a microscope? So remember he said that you'll never be able to see closer than about half the wavelength of light. That turns out to be about 100 times larger than a molecule. It's smaller than a cell, but larger than a molecule. So how do we get past this limit which is so fundamental that it's literally engraved in stone in Jena? There's a loophole and it's a pretty obvious loophole when you think about it. If you took a sharp probe that, for example, a glass fiber is the way I developed it, and then you pull that like a piece of taffy the way van Leeuwenhoek did. Then you coat the sides with an opaque material, but you leave the end so there's a little hole there. That hole is much smaller than the wavelength of light. If you put that right up against the specimen, then the light that comes out of that hole has to be illuminating an area that's smaller than the wavelength of light. You can then drive that around point by point and start to generate an image where the resolution is given by the size of that hole. Instead of by the wavelength of light. That's called near-field scanning optical microscopy. That's what I did in grad school. That's what I did when I was at Bell Labs before I dropped off the face of the earth. And so this is some of my data from that era. So here is at that Abbe limit what the cytoskeleton or some of the proteins that make up the inside of a cell looks like. But if instead you look at it with the near-field microscope, you can see all of the structure. This was the first example of a super-resolution optical microscope that could actually beat the Abbe limit. Now there are at least 4 different technologies that act as super-resolution microscopes. One of them, as Jerry mentioned, was the one that Harold and I developed in his living room a few years ago. That one is basically commercialized by almost all major microscope manufacturers today. You could also use the near-field microscope, though, not just to look at cells. It was unique in that it could be looked at anything. Here I was able to actually write bits in a data storage medium. These are near-field bits versus, versus Abbe-like diffraction-limited bits written and then read out back in the early '90s with my near-field microscope. As I said, there's super resolution is a whole other topic I don't have time to get into today. Let's continue on with where microscopy and telescopes continue to bounce ideas off of one another. One is, so now today instead of the Hale Telescope, which was that workhorse I described, had a 5-meter mirror to it. You couldn't build a bigger mirror than that because you ran into the same problem that they ran into with the big refractors, is that gravity would cause the mirror to sag. If you made anything bigger. So what they came up with was the idea of not making the mirror in one piece, but of making it in these hexagonal segments that they'd fit together like a jigsaw puzzle. So 36 segments in this Keck telescope. This is now a 10-meter telescope, or twice as big as the Hale telescope. This is a state-of-the-art telescope today, probably the second biggest telescope in the world currently. But even then, you still have to get all of these segments to have exactly the right shape relative to one another. So they would have all these little actuators underneath every mirror that would tip and tilt and move every mirror to get them so that they would line up. This is an early example of what's called today adaptive optics, where you have the mirror change shape to be able to correct for anything that's going wrong. And so with that, they were able to correct for both the gravity effect, for any like temperature changes that would cause the thing to warp in shape and so forth. So with the adaptive optics off, this is looking at little cells of convection on the surface of the sun. And with the adaptive optics on, they could do this. But this was still fairly slow. This was done a couple times a second, which is all they needed for gravity and temperature. But adaptive optics, as you'll see, goes much farther than that. So the biggest problem with telescopes was that once you have a lens or a mirror that's bigger than about 8 inches, as Huygens had back in the 18th century, you really can't do any better in terms of angular resolution, not because of the theoretical limits of what a telescope is capable of, but because of the atmosphere. So the light from the stars travels trillions of trillions of miles to get to the Earth. And it's a perfect flat wave front, a perfect image of the star until it gets the last 100 miles out of the trillion it went, it gets completely screwed by the atmosphere. And so the wave front gets distorted, so the light rays coming in, instead of converging to a point on the camera, all go to different points on the camera. And your star, instead of looking like a star looks like a big snarly mess. You know, the twinkle twinkle little star, right? The stars twinkle and that's what you're seeing. It's hell for astronomy. It means that an 8-inch telescope will perform from a resolution perspective as good as that Keck scope. So what do we do about it? Well, one solution is say forget the atmosphere, let's go above the atmosphere. Okay? Put our telescope in space like the Hubble Space Telescope. Telescope. And so this telescope has only a quarter of the diameter of the Keck scope, but nevertheless, because it's above the atmosphere and it's even half, it's half the size of the Hale Telescope, but because of these aberrations due to the atmosphere, this is the best image the Hale ever took of the Crab Nebula. This is what the Hubble can take of the Crab Nebula. So that shows you what happens if you can get rid of the aberrations caused by the atmosphere. And Hubble, like many other things that NASA has done, and this is, if nothing, an advertisement for NASA, this talk. But the Hubble has, again, revolutionized astronomy at all sorts of different levels. And this is some of the beautiful— and again, go to the website. Every day there's new stuff coming off that shows exactly what this thing can do. There's a similar problem, though, in microscopy because just as a microscope is designed, again, to get all of these light rays to converge to one point to get a perfect focus or an abbe perfect focus. But the problem is that it can't be designed— it doesn't know what it's going to look at, what the biological specimen is going to be. So it can only be designed for basically water or air or something like that. Biological samples aren't just water. If they were, we wouldn't be alive. We have to have this heterogeneity of cells and proteins and lipids and all sorts of things. All of that messes up the light rays just like the atmosphere messes up the light rays in astronomy because it causes the light rays to bend like this and they don't come to a single point and you get crummy resolution. What it means is that even if your microscope on a perfect sample would give you diffraction-limited A resolution. When you start popping it into any real sample you're interested in, it won't. How do we deal with that problem? I'm going to go on a little tour, an offshoot here instead. One approach would be to do what the equivalent of the space telescope. The space telescope, you just get rid of the problem by getting rid of the atmosphere by going above it. The way you do that in microscopy is saying, If I want to look deep in a tissue, why don't I just get rid of the tissue that's above the level I want to look? So you start mechanically sectioning, doot, doot, doot, doot, doot. You look, you section, you look, you section. That can be done optically and there's efforts here to look at that. That's called histology basically. People have been doing this for hundreds of years, but here we have very fast automated ways of doing that. This is doing it not with an optical microscope, but with my buddy Harold.— he did this with an electron microscope where they need to take really, really thin slices every time to be able to go through it. So what you're looking here is a section of the optic lobe of the fly brain which processes the visual information from a fruit fly. And what he has is an electron microscope that images one plane at a time and then uses a charged atomic beam of little atoms that whack against the surface that you're imaging and basically sandblast just a couple molecules thick off that surface. And then you re-image, re-image, re-image, re-image. And then that creates this pattern that you saw there where you can go plane by plane through the whole thing at ridiculously high resolution, about 5 nanometer resolution in X, Y, and Z. And we have these microscopes are right behind us here working as we speak, very slowly carving their way through these things so you can find out all the connections in the brain at this sort of nanometric level. So it's nice to have a space telescope if you can afford it. It's hard to service, as we learned with the Hubble when it had errors to begin with and we had to send a space shuttle to fix it. And the problem with mechanically slicing through your specimen is it's not going to be alive by the time you're done or as soon as you're starting. So it would be nice to have a solution to do ground-based telescopes at good resolution and to be able to do live specimens at diffraction-limited resolution. And so, like I said, the problem is that the atmosphere makes the Keck without any corrective measures have resolution as good as Huygens had. Couple hundred years ago. But the solution is, I mentioned this adaptive optics to change the shape of the mirror in the Keck. What they can do is they can, there's a secondary mirror in the Keck, a smaller mirror. They can put very fast actuators on that, hundreds of them. And what they do is they take the light that comes into the telescope, they bounce it off that mirror, which can change its shape so fast, about once every millisecond or 1,000 times a second. And then they look at the light as it's bounced back off with a special sensor which can sense exactly how scrambled the light wave is. That they then go in a computer that very quickly changes the shape to put an opposite curvature on this mirror to cancel these wavefront aberrations so that you get a perfect focus again. So basically you're correcting continuously what the atmosphere is doing to mess up your image, you're continuously correcting. And you get this back now to this. From signal, you go from this little bump here to signal that strong there. So it's really revolutionary, the adaptive optics, to be able to do this. This came online. You need very fast computers. You need fast actuators. This is a modern technology. It only started to come online in the '90s. So this is Neptune as seen on the Keck telescope at 10 meters without adaptive optics, and this is now with the adaptive optics. In fact, the Keck telescope, because it's physically bigger than the Hubble, in the infrared where it can do this fast enough, it's— you can see this is the Hubble view of Uranus. This is the Keck view of Uranus. So now we're really getting diffraction-limited performance out of the Keck telescope with its bigger diameter. And now ground-based telescopes can beat the space-based telescopes unless somebody put up a massive space telescope. So can we apply adaptive optics though to living organisms so that we don't have to mechanically slice them? So again, here's the idea here. One of the tricks the astronomers use is in order for this sensor to measure the distortions It needs a bright point source to look at. If you have a bright star nearby, that's good enough, but there's not enough bright stars close by to many things you want to look at for that to work. So the astronomers came up with the idea of shooting a laser high into the stratosphere where there's a layer of sodium atoms, and they can actually get those sodium atoms to glow to create an artificial star called a guide star. And then they can use that light back with their sensor. Sensor to do their correction. It's also possible— this is data we've done just probably in the last few weeks— in transparent organisms to use that idea with the same type of sensor to be able to do an adaptive optic correction. This is a model organism called a zebrafish. It's about a centimeter long here. What you can do is you can use a special type of laser called a two-photon laser. To be able to create just a little glowing point at one point in the brain of the fish. Then you can put that through your sensor similar to this sensor, build a circuit just like this, and then you can start to image all the way through the brain of the zebrafish, a live zebrafish now. This is what you would see without adaptive optics, and this is what you see when you put on the adaptive optics. Then you can start to change the focus and start to build plane by plane without actually physically sectioning the organism, start to figure out the 3D architecture inside of the brain in a living fish. But the bigger problem is that works well for fish because a fish is— this particular fish is basically transparent. But the main systems that are studied at Janelia are the fruit fly and the mouse. Fruit flies over there, mice over there. And it turns out that these organisms their brains are very scattering. So basically if you look at them under a microscope or you take a slice like you would on a ham slicer, they look like tofu, okay? So trying to image tofu through tofu and see the neurons in tofu is not an easy thing to do. Much harder task than being able to do this with a zebrafish. So it's like trying to do adaptive optics on a star when the star is hidden behind a bunch of clouds? Oops. That's the astronomers haven't figured out yet. So we had to come up with something new. So the idea was basically to go back to the drawing board and say, if the problem is that each light ray is getting bent to not come to the desired point, why not turn off all the rays except for one, measure how it's getting bent, and then tilt that ray as it's going in to swing it back to the desired point. We do this one by one with the rays. So that was an idea that I and my postdoc, Na Ji, developed, and now has taken this to a much finer level of sophistication since then. But this is one of the results from a year or two ago where she's looking in the live brain of a mouse where you cut a little window in the skull, glue a little glass cover there, and then image through that with this two-photon microscope. This is the top view and the side views going through a cubical volume of living brain tissue showing the improvement with the adaptive optics. This is the mouse here underneath the adaptive optic microscope. That's pretty much the end of my talk. I'll just say a little bit about where things are headed. In astronomy, this is the next generation. In the U.S., the Gordon Moore Foundation is funding funding what's called the 30-meter telescope. So in other words, in diameter, 3 times bigger than the Keck. This is a person here, okay? This is the telescope mirror. The Europeans are even more crazy. They're trying to, they're going to be developing the European Extremely Large Telescope. This is a person here. This is now a 40-meter mirror, okay? Again, the only way these things will work is with adaptive optics. Otherwise they won't be any good. The adaptive optics has to be extremely sophisticated. I'll say more about that in a minute. In terms of space telescopes, there's all sorts of space telescopes out there right now. The successor to the Hubble though is the James Webb Space Telescope, which would have about a 3 times bigger mirror. It's meant to look in the red and the infrared at very distant points in space. So it has to run close to absolute zero. With no influence. The Hubble is so close to Earth's orbit, in low Earth orbit so the shuttle can service it, that all the light from the Earth and everything can give problems to the Hubble and can generate heat on the Hubble because you're always in the sun. This telescope is going to go out beyond the moon in a gravitationally stable point beyond there and just kind of dwell there, be a few degrees above absolute zero and take images in the infrared that will be the best ever. It's slated for 2018. It almost got killed in the budget last year. It may yet get killed in the budget because it's very much over budget. But if it ends up ever actually making it to space and succeeding, it'll be revolutionary just as the Hubble was. These instruments will be revolutionary too, but they need that adaptive optics. And more so the problem, one problem which has not yet been addressed well enough in either astronomy or microscopy with adaptive optics is when you make a correction with adaptive optics, it's only good in astronomy over a very small patch of sky. Because atmosphere that's flickering over in this view is totally different than the atmosphere here. So you can't look over here when you're, so it's like looking through a soda straw at a very small patch of sky. Sky. Same thing when you apply adaptive optics in microscopy. As the samples get more and more complex and more and more scattering, the field over which a correction gets smaller and smaller. Astronomers have a plan, and they've started to implement that plan to be able to handle that. That's this guy here, if I can get him to go. Yes. So they've done something where they use multiple deformable mirrors. With multiple guide stars across parts of the sky to greatly increase the field of view they can look over. This is the initial demonstration on a 7-meter telescope in Chile, where they can go over, over about 200 arcseconds, or about a factor of 10 bigger than they were able to go before, and then zooming in. So these telescopes will need this multi-conjugate adaptive optics to be able to perform to spec. So they are really sticking their necks out in many different technological directions at once with these scopes. But if they succeed, we'll see the universe like you've never seen it before. We'll be able to actually image Earth-like planets around other stars. We'll be able to do all sorts of imaging on exoplanets and many, many other things at the reaches of the universe. So microscopy is likewise becoming exploding. When I got into the field of microscopy in the early '80s, I pretty much knew everybody. Now I would say I know about 1% of everybody doing it. It's getting bigger and bigger all the time. My own personal group has shifted focus from being able to do super resolution like I discussed into being able to look at the dynamics of living cells in 3D, which I think is the real frontier for optical microscopy going forward. So here you see some examples, recent examples. This is like 2-week-old data of a cell crawling on a coverslip at very high speed. We're looking at one volume through that cell, which consists of 200 2-dimensional planes every second. This is looking at chromosomes dividing in a cell. This is looking at these microtubule endpoints, which are— the microtubules are these long protein filaments that pull pull these chromosomes apart and they start from these two poles and then they start to yank to do what's happening up there. And so now we have tools by using extremely thin sheets of light to do what's called optical sectioning instead of any kind of mechanical sectioning to be able to do this extremely rapidly and without killing the cell. And this is going to, I believe, be a really big deal for cell biology and for looking at small embryos. And so with that, I'll end. I'll just say that it's a real privilege to be able to work in this field and to be part of this 400-year history of so much innovation. And it's done so much to make us understand exactly who we are and where we are. And we can all be very proud of what has happened over that period. And it's a real privilege, too, to be able to do this work here at Janelia. That's one of the reasons that I left science. Bell Labs, when I was there, was to me a heaven of work. But after the Bell System was broken up, then the whole monetary foundation that was underpinning it crumbled and the place had to go away. So once I left that, I wasn't interested in the traditional academic job because there's just too many strings attached and too many things that slow you down. It's just not for me. And I really felt lost until I heard about Janelia. And Janelia to me was like the rebirth of Bell Labs and the ability to be able to just work all the time, not worry about grants or publishing or teaching or whatever, but be able to behave the way that Hooke behaved and Van Leeuwenhoek and Herschel and those guys and spend all of my days just at the bench doing stuff. That is exceptionally rare in modern science, and I'm really fortunate to be able to do that at Janelia. So with that, thank you very much and happy holidays. If anybody has any questions, I'd be glad to take them.—

was that some theory?

The achromatic duplex element. Was that somebody using theoretical physics to come up with that design or were they just playing around with elements? So a number of people before him understood that basically they took a water lens, basically a bag of water, and they took a regular lens and they realized that they could kind of improve this chromatic problem by doing that. And then he just empirically tried different lenses, this lawyer, and eventually came up with a design that worked pretty well. It wasn't until Lister that it became calculated and understood exactly which glasses, which shapes would do it precisely. But for 50 years before that, it was all just empirical. At places like the Wilson Observatory, they're experimenting with something, I think it's called interferometry. In which they take different kind of samples. They assemble maybe 10 or 15 different mirrors, and that gives them the equivalent resolution of a mirror that's as wide as all of those samples are from each other. Is there an equivalent to that in microscopy? Absolutely, yes. So basically, like I said, your resolution in either case depends on the size of the pupil, but the pupil doesn't have to be continuous. So just like this very long baseline interferometry you're referring to. They're able to stitch together multiple telescopes to make a telescope that effectively has a bigger thing. But you lose— you don't get any more light collection, right? Because you have the gaps. So in microscopy, you can play the same games and start to use multiple elements to be able to come from different directions and increase the spatial resolution. So one simple expedient is instead of having one objective is put another one exactly opposite it. So they both come together at the same focus. That greatly improves the axial resolution and is also the basis for some— we developed the PALM microscope, Harold and I, for the super resolution. Harold's first thing when he came to Janelia was he invented the IPALM microscope that used these opposed objectives to get resolution better than even the PALM microscope. Yes. Okay, this might sound like a silly question, but I'm just curious. When you're in the microscopy, when you're trying to delve into the structure of the brain of a dead animal, it's great, it doesn't move, but I mean, a live animal is breathing. Right. How do you make up for the fact that there's movement all the time? Right, right. Right now with the state of the adaptive optics we're doing, we're looking at corrections in live animals where the movement— in a zebrafish, the movement is small enough because it's immobilized, and the adaptive optic correction covers a big enough field to deal with any movement of the fish. For the brain of a living mouse, it depends on how deep you are. If you're close enough to the surface, you give it a head mount, okay, and so if you're only a few hundred microns deep, it's fine, but if you went way deep, because the brain is jello-y, then the— particularly the blood is the thing. The pulsing of the blood can start to actually mess up your adaptive optic correction. So basically, you need to find ways of doing faster adaptive optic corrections and be able to possibly sync it to the motions that are going on. So you kind of stroboscopically do the thing. These are all the technologies that Na Ji is developing in her lab. I have two questions. The first one is, you referred to a website. Is that this website? No, I don't know which one I referred to. Oh, you were talking about all the news. Oh, NASA. Yeah, I don't have an exact link here or anything, but just Google Cassini orbiter, Google, you know, Hubble telescope, and you know, it'll be, you'll get right there. Okay, and the other question I have is, I have always been confused by the guide star concept. Can you explain that a little bit? Yeah, sure. So again, to really understand it, you have to understand how the sensor works as to why you need it. But basically, you need the way sensor that's used to figure out how the atmosphere is screwing things up assumes that it's looking at a point. And so you need to have something which, on the scale of the resolution of the telescope, looks like a point. And the way, again, that they are able to do that is by shining a laser that goes up the stratosphere. And there's a particular layer in the stratosphere which has sodium atoms in it, and only that layer and nothing else. So therefore, when the laser hits that, it excites these atoms so they glow yellow like a streetlight. You know, the yellow streetlights have sodium in them. But now you don't see the whole laser because there's no sodium atoms all the way up. You only see when the laser penetrates that thin layer of sodium atoms. So you get a tiny little spot of light, okay? And then that spot of light can be very bright compared to your normal stars in the sky. And so you get a lot of light, which is what you want, very accurate corrective measurement with that device. So I could go into the details of the device, but you probably don't want me to. Yeah. Thank you for describing a very interesting history of the making of tools to enhance our understanding of the universe around us. We have a number of degrees of freedom still to go. What do you see as the future of toolmaking in this business to enhance our understanding? Oh goodness, there's so many directions that toolmaking can go. I mean, again, for me it's a matter of It's sad because you always have to choose and focus on one or two, but there's so many different— Janelia is, in my opinion, dominantly a tool-producing facility. Most of the people here are working on one tool or another, which in the end will likely have an application in being able to understand the brain. But there's tools like Harold's tool of the focused ion beam milling for the There's tools in terms of being able to study behavior in flies and mice so that you're able to give them tasks to push levers or whatever. And then the other tools look in the brain and look at how things— there are tools in terms of developing different types of fluorescent indicators which the animal expresses naturally. And then those fluorescent indicators light up when the brain activity is going on. And so you couple that to a tool like the imaging tools that Naz developed. And you couple that to the behavioral tools that people like Karl and Ala here developed. And so there's kind of a bigger— a bunch of tools come together to start to answer these questions. One of the biggest tools is the tool that guy developed, Jerry. The ability to do very specific manipulations of the fruit fly's genetics so that you can get specific neurons to either light up in different colors or you can silence certain neurons. Basically try to reverse engineer the brain by turning off connections, doing things of that sort. And tools in microscopy and astronomy are just, you know, you never know what's going to come. I mean, like the PALM microscope, I mean it's a completely new principle from what anybody else was doing at the time. But it was enabled by— the other thing about tools is tools grow new tools. So like the PALM microscope basically came out of— well, first off, in the mid-'90s, people figured out how to use a gene from a jellyfish that they could splice into any organism to get it to glow called GFP. That was a new tool that transformed cell biology. Then in the early 2000s, people figured out a form of GFP that doesn't glow until you shine a different kind of light on it, and then it would turn on. When Harold and I heard about that in 2005, we realized that if you could take that tool, that new tool, and then shine the light on to turn on just a few molecules at a time, you could find the centers of emission of those molecules and turn on more. And make an optical microscope with molecular resolution. So that was our PALM microscope, a tool that was enabled by the new tool of these photoactivated fluorescent proteins. And when those people developed that tool, they had no idea that it could possibly ever be applied in that way. That's the wonderful thing about science, how combinatoric it is. We just have— you can see this sort of exponential growth of our abilities to be able to explore the universe at all levels because the tools get better and better and better as we go.

And that's the history of science. Maybe that's a good grand notion to end on.