 Thank you very much for inviting me. This is, I guess, my second visit in less than a year. So I'm really enjoying meeting several people here. It's good to see you all. It's an interesting site. It's hard to tell you apart. It sort of answers the age-old question of why zebras have stripes. The last time I was here, I spoke about something completely different, which was on the topic of how insects build structures, termite architecture, and so on. But this is more what my lab focuses on, on the questions of how insects fly. Now, this has been a question that has driven me for at least 20 years. I started out with the idea that I'd try and answer a few questions that I had, and then move on to something else. And I find here, now, 20 years later, still just as fascinated or more fascinated with this question than I ever was. And that's in large part because in science, the more you find out, the more you know that you didn't know. And this is something that just helps you in the process of finding questions. Now, I'm not going to talk about this question from not cover all aspects of this question. It's just too vast a topic. So what I've left out, and I'll already tell you at the very beginning of the talk, is I've left out aerodynamics from this. We will brush upon some of the principles, but I will not talk about a bulk of the work that we've done on how insects generate flight forces. This is maybe a topic for some other time. And I want instead to focus, I also will not talk for the most part about the nervous system. Although again, I'll brush up on that and I can't but help at least touch upon that topic. But the specific thing that I want to talk about today is the sub question here, which is how insects deal with the challenge of miniaturization. But let me start by showing you a movie. I'll show you two quick movies. And this is an insect which all of you have seen maybe as early as this morning. This is a house fly. I think all of you have seen this fly and I'm sure you've seen it many times. I'm sure what you're going to see is also something that you've seen many times. The only difference is that this is filmed. The movie that you're about to see is filmed with a high speed camera. So this is filmed at almost 4,000 frames a second. Your normal camera operates at something like 25 to 30 frames a second. So that's the only difference. But as you will see, when you have a high speed camera, it's almost like having a microscope in time. So you are able to see things that you wouldn't otherwise be able to see. And what the two behaviors that I've chosen to show you are both very mundane behaviors. One involves looking at the fly as it takes off. You've seen that all the time. And another involves the fly as it lands. And I want you to pay attention to just how exquisite both of these things are. Just some quick facts. This is, as I said, a house flight. It's called Muscat Domestika. And it flaps its wings between 200 to 250 times a second. That's about 4 milliseconds per stroke. Just keep that number in mind as we watch these two movies. So here's one that you can see. What I want you to notice is that as soon as it has left the substrate, it is already pretty stable. And it's doing a lot of very interesting things with the two wings, as it is taking off. But I think all of us will agree that this fly knows what it's doing. It's stable. It is not struggling to fly. It's not likely to fall off the air. Now here's another one, which is even more fascinating. And I should tell you that this was a completely accidental movie. We did not intend to get this. This was a complete stroke of luck that we got this movie. Now what you will see here is a fly coming in and landing on this glass vial. Now if I were to tell you that we carefully engineered this behavior to occur here at this time, I think you should be skeptical. Because firstly, flies don't do what you want them to do. They have their own mind. And it's a fairly active mind. I can assure you that. But also because to get a movie like this does require some level of planning and sophistication. And the reason for that is, how many of you do photography? Few. So if you work with cameras, and if you work especially with these things called macro lenses, which allow you to look at very small things, then you know that the trade-off with trying to look at very small things is that you don't have depth of field. So you can only sort of look at things that are very in a very small region of space. And so to get a fly to come and sit in that small region of space is quite a challenge. So I should say, so we were lucky, the fly sort of, we were thinking of looking at a take-off behavior just as the one you saw before. But the fly took off. I was slow to hit the trigger. So what you do normally is, these are high-speed cameras. So you want, you're sitting there with a trigger. And the cameras are taking films continuously. And what you can decide with the trigger is what part of the pre- and post-trigger time points you want to capture. So you can say, for instance, I want to capture a second before and a second after the trigger. So here's the movie. Here comes the fly. And I want you to notice that it's front legs already up. So this fly knows it's going to land. And it's going to now pitch up. This is all happening with the aid of these two wings. Slows down and lands. And I was lucky that I hit the trigger slightly late. So I missed the take-off part, but I got the landing part quite by accident. Now what I showed you are examples of what we, in biology, call behavior. This is what an animal does, but with a slight difference. The things we call behaviors are typically things that you can elicit repeatedly. So I can film take-off flight any number of times. And I know that I can get a fly to take off. I can either take an object close to it and it will fly away or something like that. But it will keep doing this again and again. And you treat this very much as a statistical entity. So you try to control the conditions under which you elicit this behavior. And you repeatedly film this behavior. And then you treat it very much as a statistical entity, meaning that you're looking at an ensemble, all of which more or less behave the same way. And then you try to look at what are the generalities of all these behaviors. So what I've just told you is that part of the challenge is to see a behavior out there in the wide and then bring it to the lab and try to reconstitute it in the lab under controlled conditions. And once you've done that, then a whole set of questions open up. So this part is just about where you find those behaviors. Flies are looking for mates. They're looking for food. They're guarding territories. They're navigating. They're navigating in short or long distance. All of these lead to behaviors that relate to flight. And once you've found a behavior like that, then a whole stream of questions open up. You can ask, what are the sensory cues? For instance, I loomed an object in front of the fly. The fly got, let's say, scared to use an anthropomorphic term. And it took off. So vision was involved. Maybe I gave it a puff of air. Maybe that causes flies to take off. Maybe I make a sound. Maybe that causes flies to take off. All of these are things that you would term as sensory cues. So when I say sensory feedback, it's actually multiple arrows going in. It's not just one arrow because there are multiple sensors that are feeding into the central nervous system, which then sort of combines all this. And there's an absolutely wonderful world of computational neuroscience out there which tries to address how exactly these computations occur. And all of these computations occur. They are then converged down to a set of neurons called the motor neurons. These motor neurons are those neurons which connect to the muscles. And these muscles are then sort of there's a lot of them. And they perform a concert of actions which then causes the behavior to occur. But the muscles can only ever contract. You all know that. And when they contract, they pull on body parts. We have an internal skeleton. They pull on our bones and joints. Insects have an external skeleton, an exoskeleton. So the muscles are pulling on the thorax. And we're going to see a lot of that in the later half of the talk. So they pull on this region, which we call the musculoskeletal system. And that causes thoracic oscillations. The thorax will then sort of oscillate. And because of those oscillations, which are transduced by an extraordinarily complex wing hinge, which again I'm going to show you today, into motions of the wings. So the wings then move. And when they move, they interact with the external fluid medium, which is air, which then causes locomotive forces, torques around the body. And that is what eventually leads to what we see as behavior. So the aerodynamics of insect flight is one aspect of the question. The musculoskeletal system is another aspect of the question. The sensory motor feedback, how the nervous system works is another aspect. These are entire fields in and of themselves. And then, of course, there's the ecology and evolution part of this problem. And all of these sort of combine when you're studying insect flight, which is why it is such a fascinating topic. When I started out, this is all I cared about. And in the process of answering questions there, I've discovered all of these other things. And I think I'll remain employed for a while. So those are all the different projects that are ongoing in the lab. There's quite a few of them. But I'm just going to talk about a few of them. And there's some other projects, including some plant biomechanics and termite architecture and insect architecture in general. I'll be happy to talk about these if any one of you want to discuss them later. But let's start with flight. And why worry about why be so fascinated with flight? I think all of you recognize that insects are the most successful multicellular tax on the earth. They have always been and they always will be, I think. They've been around much longer than we have. I think they'll survive us. There's an estimated 6 to 10 million species. And we've barely covered 1 million being able to describe them. And I hope we get to the others before we get rid of them. They are a substantial majority of, I think, that's more like 80% of all multicellular animals. They're flying insects range from size scale spanning three orders of magnitude. This is a topic of discussion today. Think about it. It's absolutely spectacular. We are roughly a meter in size. We're thinking of something that would be about a kilometer ranging from our size to that of a kilometer in size. I mean, on the order of. That's the kind of size range we're talking about. They occupy a vast variety of ecological niches. You'll find them in Antarctica. You'll find them on the tops of Himalayas. You'll find them in water, all sorts of places. And they have been around since at least 400 million years. So if we just focus now on all these things on the size scale, then there's some very interesting things to think about. The first is, so this here is the fossil of the largest insect that we know of. And this is a dragonfly, which was from the Carboniferous period, about 300 million years ago, which had a wingspan. So the distance between this and that was about 65 centimeters. That is a scary insect. But this is extinct now. Of the extant insects, you have this one probably takes the price for the largest insect. This is Queen Alexandra's bird wing. We have something that's pretty close to that size. In the Western Ghats, you'll see something called a southern bird wing, which is about as big as this. And that wingspan is about 1 foot, 30 centimeters, large insect. Now at the other end of the spectrum is this remarkable animal, which was described by Polilove recently, which is a type of wasp called the mega-fragma mymarapen. And it is about a little less than 200 microns in size. That, for comparison, is a cell of a paramesian. These are both two scale. And you can see that this animal is smaller than some single-celled animals. Yet you can see that it has a nice antenna. It has eyes. I'll show you these eyes in a different insect, but close in. It has wings. These wings are more feathery in structure. I shouldn't brush-like in structure, not feathery. But brush-like in structure, not paddle-like. But they're so small, it doesn't matter. They both act the same way. And they have a functioning nervous system, and they're doing very interesting things, almost like any other insect that you see. Here's one that we are working with. It's a trichogramma wasp. It's not as small as the one that I just showed you, but it's about 400 microns in size, still less than half a millimeter. And literally, if you take a pencil and you put a dot on a white sheet of paper, that's how small it looks. And yet, look, it looks very much like a normal insect. This one doesn't have those feathery structures. You have these kind of vessels coming out, but its wings are paddle-like. It has eyes. These are the compound eyes, which are shown here in a close-up image. And you can see the individual units that are also blown up here. So, and these fly. As you can see in this video, this is something. It's very difficult to get these videos, because you're focusing on a very tiny area with a high-speed camera. And so you're not almost always you're not, you don't have sufficient light. If you try to put too much light, the insects just dehydrate. So it's really kind of difficult. But we managed to get this one. This is Abhin and Akash, Burdan and Abhin Ghosh. And you can see immediately that they do this curious kind of clap of the wings. I'll show this to you again. So the question that really is fascinating to us at this point in time, having studied a little bit about how they generate like forces and aerodynamics and so on, is how do you get a nervous system? How do you pack it into something that small? And how is it able to conduct its business given that it's been packed in such small space? And we're going to try and address that just to sort of general principles of it. But before we do that, just for a minute, think from the perspective of a small insect. What are the challenges that this insect is going to face? One of the problems that insects face when they are small is that they have to, I'll show you in a little bit, that insects, as you reduce their size, they have to flap faster. So a lot of these high winged frequencies that I just showed you come from the fact that they are miniaturized. And there's a physical reason for that. So if you need to flap faster, then you need to acquire information faster. So the sensory system has to act with much higher temporal resolution than we are sluggish in comparison. If I lose my footing and I stumble, it takes me up to a half a second to a second before I'll actually hit the ground. Insects don't have that luxury. So they need to act much faster. The motor system commensurately needs to be faster and more accurate. So it's useless if only your sensors are fast and your motor system is still slow. Then, of course, you have things like energy losses. You're flapping many more times, which means that you're wasting a lot more energy because there's always some loss. Now you have to reduce those losses. And that is going to, the way to reduce them requires elastic storage mechanisms. Insects have one of the best known rubber in natural world or in any world. It's about 97% efficient. It's called Resilent. It's an extraordinary biological material. Of course, they lose water. Those must be compensated. There's many, many more questions like that. Why do they need to flap faster? And that's a very simple explanation for that, which So the equation for flight forces is this. It's half. There's a coefficient of, just think of it as a multiplication factor times the density of air. Let's not worry about that. The velocity of the wing, the square of it, and the area of the wing. And so if you break down the velocity of the wing into the angular extent of the wing, that's this angle phi, N, which is the frequency of the wing and the length of the wing. So if you work through this now, you find that the flight forces go as a cube of the wing length. But because the chord length, so this is the wing length and that's the chord length, that chord length is also a function of wing length. So it actually goes as the fourth power of the wing length. So you can increase flight forces as the fourth power, but your mass is only increasing as a cubic power of r, which means that you have to do something in order to maintain those flight forces. And that can only be done by increasing the frequency. And sure enough, when you look at insects, as you decrease in size, the wing bead frequency increases. And this is shown here in two ways. One is in terms of decrease in size with body mass as a metric, and here is just size as a metric. And in both cases, what you find is that as body mass decreases or as wing length decreases, you find that the wing bead frequency increases. These are log-log plots, so you can see that they actually increase almost by an order of magnitude. And they can go up to almost 1,000 hertz. Mosquitoes have been recently shown to operate something like 750 hertz. Mages go up to 1,000 reportedly. So this is remarkable. Let's just put that in perspective. Your eye blink is about 150 milliseconds, about, let's say, somewhere between 150 to 200 milliseconds. In that time, if fly has completed something like 50 wing strokes, there should be no surprise that you can't spot them. They are far too fast for us. So the question is, how are they able to do this? If you've learned about muscles, you know that for muscles to twitch or for muscles to contract, they need a stimulus from a neuron, the motor neuron. But can the nervous system operate this fast? The nervous system comes up with its own set of constraints. Typical action potential is on the order of one or two milliseconds. So how are you going to pack several action potentials which are required for a wing stroke into that short time frame? Now the answer to that is very, very interesting. And I'm going to just take you through that. I hope I do justice. So certain kinds of insects, especially those that have miniaturized. So a few orders of diptera, which is flies, mosquitoes, et cetera, colyoptera, beetles, then hamenoptera, bees, wasps, et cetera. These have evolved a kind of muscle, which is called the myogenic muscle. Then the myogenic muscle is a muscle that doesn't require a constant neural input to contract. All you have to do is take the muscle. In an in vivo case, you extract the muscle, keep it in physiological solution. And if you stretch it, then with some delay, it contracts. All it requires is a stretch. You can cut the neuron that feeds it, and this will still work. Now if you compare an insect that does not have this myogenic or asynchronous muscle form, it's called asynchronous, because it doesn't require a one to one neural stimulation. So if you compare an insect, this is a larger insect, flaps much slower than this beetle. You can see that the wing movement of the beetle is much faster than that of a locus. But the underlying nervous system is not that much faster. So this thing has reached its limit, more or less. Whereas here, there is a one to one correspondence between the nervous system and the wing flapping. Here there isn't. This is why it's called asynchronous. And if you look at the physiological properties, then you see this kind of delayed stretch activation occurring. So here's locus, and here is beetle. So this is the stress on the muscle, the force per unit area with which you stretch it. And then this is the strain on it. And if you take a small part of it, you actually see what's going on. And you can see that in this case, whereas when the strain stops, the muscle sort of comes back a little bit. Here it continues to extend to contract. Now you take these muscles. Remember this stretch activation property, because I'm going to come back to it many times. Now you take these muscles, and you pack them in the following way. This is the thorax of the insect. That's where the head is. That's where the abdomen is. And the green muscles here are the ones that we call the docile longitudinal muscles, because they go sort of longitudinally. And the, I don't know what this color is, peach muscles. The ones that go from top to bottom, and they call it docile ventral muscles. So these are an antagonistic set of muscles, just as we have flexors and extensors. So the flexors do the opposite of what the extensor does. Similarly, you have these docile longitudinal muscles and the docile ventral muscles. So now if these contract, the thorax is going to be pulled in. And if the thorax is pulled in, these docile ventral muscles are going to be extended. Now I told you that the stretch activation property means that if this is activated after a short time, it's going to contract. And now this will be extended. And then this will contract, and this will be extended. And this way, just through the physics of the musculoskeletal system, you can get many strokes for each wing beat, or for each neural stimulation. Is that clear? And this allows insects to, by an order of magnitude, increase the wing beat frequency. And you see this has happened several times. So this is the phylogeny of insects. So it's sort of like a family tree of insects. So you see all the insects here. These are dragonflies up there. Locusts are right there. But if you look down here, in sort of the more recent taxa, I should hesitate to call these more recent. But they have sort of specialized. And that's where you see two things. One is a lot of miniaturization. And two is the presence of this myogenic muscle and this indirect flight muscle architecture. The indirect flight muscle architecture is just this. This is the idea that muscles don't directly connect to the wings. They connect to the thorax, the inner shell of the thorax. And they are actually moving the inner shell of the thorax. So if you combine the two, you have this combination. Allows you to raise the wing beat frequency. And that's what you're seeing here. And you see that immediately, you start seeing the hyperdiverse order. Hyperdiversity is a measure of success. Then this has allowed insects to be more successful. And it shouldn't surprise us, therefore, that we see so many beetles and flies and bees. Lepidoptera don't have this kind of myogenic. So they are a bit of an outlier. We don't know why they are this successful. But certainly wing beat frequency is not the reason why. So this is just what I told you. You have the indirect flight muscles, which contract causing wing elevation. So in the DVM contract, the dorsal ventral muscles contract. You have wing elevation. When the dorsal longitudinal muscles contract, wing depression. And this cycle causes wings to move very rapidly. Now in addition to that, in addition to these muscles, which actuate the big motions of the insect, you have many tiny muscles. And I'll talk about them later in the talk. They are called direct steering muscles. And just hold that thought. I will come to it later. So we got into this question primarily through a very talented graduate student, Tanvi Devra, who worked in my lab. And we started asking questions about how can insects be so fast and yet so precise? And they have to be precise. If they are not precise, you will not see those exquisite maneuvers that I just showed you at the beginning of the talk. Insects will wear off course. It doesn't take much for something that light to wear off course. So how can they be both fast and precise at the same time? I think most of you are aware, many of you might play cricket or other games. If you try to increase your speed, your accuracy trades off. You're not able to be as precise. But insects seem to be able to do both. So let's talk about this question a little bit. So insects have sensory systems which are, as I said, operate very fast. And they are able to tell the insect in real time how they are doing. And this is not very different from us. We have our inner ear system, which allows us to keep balance. If I perturb the inner ear system or if there's a disease there, then it causes vertigo. You'll have a chakkar and you'll fall. So similarly, insects also have something that serves the purpose. But remember that insects are moving in three dimensions. What it has is a structure that is a gyroscope. How many of you know what a gyroscope is? So gyroscope is an instrument that used to be used by captains on ships and so on to tell about whether they were keeping a steady course, or whether they were. It's very hard in the sea to know if you are too tilted or something like that. Gyroscopes allowed you to keep that sense. And these are vibrational gyroscopes. Normal gyroscopes are rotational. And what you can see is right there, you see that white structure that's moving exactly anti-phase to the wings? That structure is the gyroscope of an insect. It's called the hortier. It used to be a wing. One of the things that has happened through evolution is it's reduced in size. It's become this mechanosensory organ, which is then able to sense how fast the insect is turning in air. And so this is the hortier right there. And if you look at it closely, what you see, this is the hortier. And if you focus on these two structures here, you can see this almost corn-like appearance. Each of those bumps is a mechanosensory unit. One neuron is connected to that particular organ. It's called a Campaniform Sensory. And it's just a little bump on the cuticle, but it measures strain on the cuticle. So and the field of these are packing a lot of information about what forces the hortier is experiencing. So the way this works is that the hortier sort of moves in a plane. And if the plane of the rotation changes, then because of conservation of angular momentum, it's going to try and maintain its plane of rotation. And that causes a force on the hortier. It's called the Corioli force. And that force is what causes the hortier to sort of bend. And those bends are being detected by this sensory system, which then reports this real time to the nervous system, which then makes sense of it and says, OK, you're turning left at this rate or you're turning right at this rate. And that allows them to get the constant feedback, which allows them to maintain stability. If this feedback was slower, the insect would have a harder time maintaining stability. So these companion forms in Silly talk directly to the wing, this wing steering muscles, as also do inputs from the visual system. So somewhere here, information from vision and mechanization is being integrated. And the brain is making sense of it all to try and understand how the insect maneuvers. Now if you knock out the hortier system, then you have a problem. So here's a normal insect taking off. This is here's an insect in which both hortiers are ablated. Here's an insect in which only the left hortier is ablated. You can see it's turning in one direction. And you can see that with its wings, it's trying to actually stabilize. But it cannot because it doesn't have accurate feedback. And this is where the right hortier is ablated. And it's just fine for a little bit, and then suddenly we are out of control there. So it's a very important sense. Now what you noticed in that movie that I showed you was that the hortier was exactly anti-phase to the wing. And you can see that here. So the wings are exactly in phase. And they have to be exactly in phase. I mean, they can't even be a little off because that's going to cause imbalances that cause them to turn around. So they're exactly in phase. The hortiers are exactly out of phase. And the natural question one must ask is how can the nervous system operate this fast? This is happening in the insect that I showed you within 10 milliseconds. But it's true even in drosophila or the house flies, which are 4 milliseconds. How can the system be this fast and yet this precise? So the obvious answers that come to mind are maybe it's all newly driven. So what you have here is a wing sensory neuron that talks to the hortier motor neuron. It's in a reflex kind of a loop. And so this causes when the wing goes up, the hortier goes down. Or it causes you have the hortier sensory neuron talking to the wing motor neuron, hortier goes down, wing goes up. Or you have one neuron that tells the wings to go down, hortiers to go up. But even that is going to be slow as compared to what is needed. So the counter hypothesis that we had or done we had was that maybe this is not newly driven at all. Maybe this is a mechanical coupling. In which case you don't need the nervous system. All you need is mechanical connections that cause one to go up and the other to go down. And this is an easier hypothesis to test because all you have to do is wait for the insect to die and then work on a dead insect. So that led to a series of experiments which we call the dead bug experiment. I'll just show you a quick video. So this is a dead insect. It's not gone into rigor mortis yet, just freshly dead. And what Tanvi does is move just one wing. And watch what happens when she moves just one wing. I think it's obvious what's happening. What's happening is when you move one wing, the other wing moves. There must be a mechanical connection between these two wings. And then the hortiers move opposite. So all of this is mechanical connection. So Tanvi joined as a very keen neurobiology aspirant. And by the time she had finished her first experiment, she was sure that she was not going to be a neurobiologist, at least not in as long as she worked on this project because the whole project veered towards the side of mechanics. And so she spent the next four or five years actually very meticulously charting out where these connections were. How do they connect to each other? And how is this all this actuated? And so I'll just give you a quick summary of what she has found out. So this is the thorax. That's the front of the thorax. It's called the scutum. That's the hind of the thorax. It's called the scutulum. I've given it a gray shade. And if you look at it from the side, you can see that the scutulum has this little arm that comes right under the wing. These are horrible names. I don't know. Many years ago, I kind of ran away from these names. Partly what drove me to physics was these horrible anatomical names. So I'm sorry I'm imposing them on you. But when the right question comes along, you've got to learn them. So I'll just summarize Tanvi's work in a video. So this gives me a break while you watch the video. This movie summarizes her findings on the biomechanics of wing and halter coordination in flies. Her study shows that precise coordination between wings and halter is achieved not by the nervous system, but by mechanical connections within the thorax. A structure on the thorax, the scutulum, links the two wings. Rapid coordination of the indirect asynchronous flight muscles drive the motion of the scutulum, which causes simultaneous in-phase movement of both wings. When this link is severed, the two wings become uncoordinated, as shown in the next video. Thus, the scutular linkage is necessary for wing-wing coordination. Although the two wings are not coordinated anymore, the halter's on each side oscillate antiface to the epsilon wing. When the two wings become uncoordinated, the halter's also become mutually uncoordinated. We next showed that the wing-halter coordination is mediated by a separate mechanical linkage, the sub-epimeral ridge, which connects the wing base to the halter's base. When the sub-epimeral ridge is intact, the wings and halter's oscillate exactly antiface with respect to each other. However, when this link is lesioned, wings and halter's become uncoordinated. Notice that the halter continues to move through its full amplitude, driven by the asynchronous halter muscles. The halter on the right side, whose link remains intact as an internal control, continues to oscillate exactly antiface with the contralateral wing. Thus, the sub-epimeral linkage between the wing and the halter's system on both sides are independent of each other. If the wings and halter's are constrained to move in synchrony by mechanical linkages, how do insects achieve control of just one wing at a time? To address this question, we propose the hypothesis that there exists a clutch at the base of each wing, which can engage and disengage the wing from the mechanical linkages. When the clutch is engaged on both sides, the two wings flap together. However, when the clutch is disengaged on one side, one wing remains folded, whereas the other can flap. Apart from the clutch, the base of the wing contains a gearbox. Once the wing is engaged, the gearbox controls the amplitude of each wing. If we zoom into the base of the fly wing during active flapping, we can see the wing edge. It consists of a radial stop shown in red, a plural wing process shown in yellow, and Terral-A-C, a putative mechanical sensor and damper shown in blue. The radial stop contacts the plural wing process in four different modes, mode 0, 1, 2, and 3, as shown here. In this scanning electron microscope image, we see how the radial stop connects with the plural wing process in four different ways, from mode 0 to mode 3. Here is a video of the wing engagement at the start of flight, as the radial stop moves from mode 0 to higher modes. Notice the shift in the wing amplitude from very low to very high within a single wing stroke. Once engaged, the wing hinge shifts between the different modes, and the wing moves at high amplitudes. This is akin to the gear change operation in automobiles. During flight cessation, the wing abruptly transitions from high amplitudes to low amplitudes within a wing stroke, as seen in this video. When this happens, the radial stop moves from higher modes to mode 0. So that's what we found out. And so that's the general model that we have for the thorax. So these are all the mechanical linkages. The wing Hortier linkage, which we've given our own ugly biology name. It's called sub-epimeral ridge, because it goes under this structure called the epimeron. Doesn't matter. It's just an ugly name that we thought we should impose on biology. The indirect flight muscles that cause the wings to move together, then the Hortiers, which have their own sets of muscles. So this is, in physical terms, a coupled oscillator model. So these two systems are coupled oscillators, and they are weakly coupled. In other words, to show that, what we can show is that if you start to clip the wings, you can actually perturb the frequency, because this is a resonant system. And so what you do is you can increase the frequency. And as you increase the frequency, the Hortier follows suit. But beyond the point, the wing continues to increase the frequency, but the Hortier comes right back to its original frequency. And this is classic behavior that you see in weakly coupled oscillators. If it was strongly coupled, the Hortiers would always stay in place with the wings. And if it were uncoupled, it would be flat. Now if you think about this, this is the sort of thing that allows insects to actually cope very nicely with wing damage. So when there is wing damage, there's going to be a change in its frequency. But because the wing and Hortiers are weakly coupled, for substantial amount of wing damage, in this case, almost 40% of its wing is gone. But the Hortier continues to be out of phase accurately. So the mechanical linkages are actually a very nice way of doing this. We've gone one step ahead. And actually in the Sofila system, we've asked about questions about what's happening now in the nervous system. I told you that there is a clutch. But the implication of that is that the two clutches should be engaged synchronously. How does that happen? And what we found out, this is in collaboration with Professor Geithi Hassan and her student Sufiya Sadaf at NCBS. We found that there is this neuron called the ventral unpaired medial neuron. This is in a genetic model system of Sofila, which throws true projections, bilateral projections, into the motor neurons of the two wings on both sides. And that allows this structure to simultaneously, or this neuron, to simultaneously activate those two neurons and engage the clutch. And so if you put it together now, you have the, this is now looking from the top. You have the docile longitudinal muscles, the docile ventral muscles, the linkages, both the wing wing linkages in green and the wing Hortier linkages. And this sits on top of the neural tissue, which is shown in gray. In the last little bit, let's come to the direct flight muscles, the steering muscles, which I promised I'd tell you about. There's many of them. There's almost 16 to 18 pairs, depending on which insects you're looking at. And we needed to look at them, but they're all inside. And so it wasn't easy. And it required collaborating with Dr. Namrata Gundeya at the IISC. And a lot of very talented students, all of whom have gone on to do two nice places. And so here's, again, a video that shows you what this looks like. This is a micro CT image. Micro CT is a way of looking at structures internally. And we really need to do that now because there's a lot of action that's happening inside. So here's the radial stop and the plural wing process. I think this video is misbehaving. Sorry. I'm going to switch out of the, maybe this works better. So that's the lateral view of the wing hinge. This is widely considered one of the most complicated joints in Animal Kingdom. So that's the radial stop that connects to the wings. That's the gearbox that I showed you, the free mechanosensor and damper. This is the part that coordinates both wings. That's one sclerite. This is a plate on the cuticle. And that's controlled by these two muscles. Now let's, this is another sclerite. This has no muscles. Axillary sclerite 2. This one is axillary sclerite 3. And it has several muscles. So 3-1, 3-2, 3-3, 3-4. And that in combination looks like this. Then there's a fourth sclerite, also composing the hinge. So this is, again, five muscles that actuate this sclerite, which together then look like this. And then you have the bacillus sclerite, which is responsible for major movements of the wings. And this has three muscles. And the mutual organization of these muscles is also a very fascinating topic, which is an extremely difficult one. From a physics perspective, this is something I would call worse than a three-body problem. Because there are several muscles. And the constraints don't bring it down to less than 3 degrees of freedom. So these are all the muscles, the indirect flight muscles. And then the steering muscles, as you can see, are off to a side. And these are all involved in controlling how the wing moves. You can imagine how complex this system is. So I'm nearly at the end of my talk. So I want to summarize just a little bit about how insects, miniature insects, and what are the features of miniature insects that are to be seen. So in general, so you have antennae, have fewer segments. They have fewer scents. They are operating at a lower Reynolds number. So they perceive the air as being much more viscous than we do. And so their odor perception may be diffusion limited. This is a rather important point that I can't elaborate on right now. Eyes, again, fewer of material. Individual units of eyes cannot go smaller than a certain amount because they are limited by diffraction. It is only so small you can make a lens before it begins to be useless. And so there are smaller diameter lenses. So they can only have fewer ones. So that means that they are not very well resolved eyes. But they do have this other type of eyes called a celly, which are fine. Then you have wings that are often reduced or lost. They have typically fringed margins, sort of like this. A lot of them show that clap and fling maneuver that you saw where the wings sort of clapped together. And they have enhanced wing bead frequency. They have a few thorax. And you saw there's all these linkages within the thorax that allow them to coordinate things extremely fast. And then, of course, the flight muscles, which I spent a fair bit of time on. But I did not talk about some of these flight-related behaviors. How did this first, even though they are very tiny, they can sometimes go over large distances. An interesting story to tell you. Some years ago, we were in Panama doing some work on migratory insects. And we were catching these butterflies out on the lake gatter just outside of the Panama Canal. So you would chase these insects in boats, capture them, and you do experiments. And we found that every butterfly that we captured had a nice red bindi. It just made no sense. But we looked at it closely. Turns out it was a little mite. And the mite was exactly between the two eyes. And we thought, well, how does it get there? And I was talking the other day to somebody who's an expert on mites. And he said, that's very simple. All they do is they hang out at flowers. And when the butterfly puts its proboscis in, they just climb up the proboscis and sit at the end of the proboscis, which is exactly between the eyes. And so they're hitching rides on these migratory butterflies. And that's how they spread far and wide. So this is to summarize everything I've sort of told you. I'll just leave you with a couple of videos. So here's finally, we were able to get this behavior, the landing behavior in the lab. And this shows you how they land. You can see now that we have two views. So we can get 3D information about the wings. And what I want you to notice, I'll show this video again. I want you to notice that right here, the wings here are perfectly in control and sort of large amplitude motion. But right about here, you'll find that the amplitude goes down. And I suggest to you that this is a fly going into neutral gear before it lands. And you can see that in the plots right here, that the amplitude goes down. Here's a movie of a fly that's changing another fly. This is the territorial behavior. And watch how fascinating this is. This is happening in three dimensions. And it's happening in extraordinarily fast timescales. And you'll see this all the time if you're paying attention. So this thing entered this fellow's territory. They're guarding territories, both are males. Everything you're seeing is happening under less than half a second by my estimate. So it is sort of sumo-result to the ground. Now, if the same fly, this fly is not necessarily always aggressive. Sometimes it's very romantic. And so if the fly actually happens to catch a female instead of a male, then this is what it does. It carries this female. You can see that its amplitude is very high. The female is not even trying to fly. She's just happy to be transported. The fly will gently place her on a substrate. And immediately, when it doesn't have to carry her weight as well, the amplitude comes down, as you can see. There you saw that. It's generating less forces. These are the sorts of things we really want to get at. And everything I've told you allows us to get insights into these kind of natural behaviors. And that's really the joy of what we are doing. I mean, to be able to go out, see a fly while you're drinking a cup of tea, and just enjoy what it is doing, and know that inside of this fly is an extraordinarily complex machine. That's really the joy of this kind of work. So I want to stop there. It's all these people to thank, many more actually than these. Just broadly speaking, everyone in black are faculty members elsewhere that I collaborate with. So on the issue of musculoskeletal mechanics with Dr. Geithi Hassan, Namrata Gundeya, aerodynamics with Dr. Shunyan Deng and Bo Cheng, both in the US, on the migration studies, which I didn't talk about today, is Professor Robert Dudley and Bob Strigley. Then the ones in blue are all students who are in the lab at the moment. The ones who are in lighter blue are people who have escaped to bigger and better places. And the funding agencies, which have been extraordinarily kind. One of the nice things about work like this is that we are not curing cancer. We are not even trying to cure cancer. We are trying to just understand nature. Somebody somewhere is going to be looking at these results, and they're going to be asking, what can we do with this? That's their job. Our job is to figure out how nature works. So I'm just going to stop there. This is a picture of the lab, in a particularly thoughtful mood. Thank you very much. How about a bithorax mutant in Drosophila, bithorax mutant, which has got the haltier changed to the wings? Yeah, unfortunately, they haven't been able to actually generate any muscles in that latter wing. So the mutant is pretty sick. It's not able to do much. Can I ask one more? You said about the indirect muscle function. So how is it connected to the direct muscle? So that's a good question. So the indirect flight muscles are actuating the big motions of the wing, the large excursions. Angular excursions of the wings are because of the thoracic oscillations. The steering muscles are doing the subtle changes. So if I want to change the angle of attack, if I want to change the phase, or if I want to move the wing back sooner rather than later, then that's where the steering muscles come into play. So the importance of steering muscles is almost on a stroke-by-stroke basis. And the neurons that supply them actually operate very, very fast. Some of them even every stroke. But their specialized neurons. Now it is not true that neurons can't operate at very high frequencies. They can go up to 500, 600 hertz even. But they can't do that and generate power at the same time. For a large muscle to contract takes time. But a small muscle can contract and can receive neural input at very rapid times. But these are mechanosensory neurons. Well, they could be, for instance. It's very interesting to listen. Very exciting to tell people. And the biology department keeps fighting all the time. And the same thing when I teach biology, I keep telling what is there in the name. I can replace it with A, B, C, D, X, Y, Z, still. It makes the same sense. But the students get driven away by so many nomenclature. I had a two problem when I was studying. A little extended question. I'm not a little student. But how in the evolutionary time scale such a complex machinery, if one visualize, you start believing as if somebody would have done that. Because the insects, even they're evolving into 400 million years time. This is point number one. And second, even in terms of material, one is excellent. The dynamics, the communication, the chemistry of neural communicates. But in terms of material, one is the muscle. I'm more concerned about the thorax. The one which is the muscle is doing the thorax is constantly undergoing kind of a contractions and relaxation. So this has to be equally of a highly elastic in nature, other than the muscular protein and other. I was just curious to know how much work has been done on the insect, other than knowing that they're the chitin, they're carbide. The material part of it. So to answer your first question, which is how does a complex structure like this evolve? Actually, this is a question that's very nicely addressed by someone called Dan Eric Nielsen. When you're specifically looking at the evolution of an eye. An eye, I would argue, is about equally complex. And what he asked was how do you go from a simple photoreceptor to a complex eye? How many iterations do you need? And so this was a computational study. Richard Dawkins has written a very nice piece on it. What he showed was that it takes very few iterations. If you give it just a slight selective advantage, then you don't need too much to go from a simple photoreceptor to a very complex eye. Then I would argue very much the same for the insect wing hinge. It is an extraordinary complex structure, but 400 million years is a long, long time. And it's staggering how long this time is. We humans have not even occupied an eye blink of that time. And yet look at us. I mean, look at the effect we've had on our environment. So we're talking of something that has lived through all kinds of environmental conditions through the asteroids hitting Earth and so on and so forth. I mean, these are strong selections. And of course, something complex can evolve from that. Coming to your second question, which is about the materials of the, yes, a lot of work has been done on the material. We know a fair bit about resilient, for instance. There's still a lot left to be known. We know that there are such structures. So the chitin itself is a complex of many different proteins and carbohydrates. And it has structural properties that often are geometrically or the sort of, so you can have a thin or a thick cuticle. And the properties are different. It's all fairly, I wouldn't say we have studied this question in detail in relation to the physics of it. But I think the biochemistry is very well worked out. So we know a fair bit about the materials, for instance, as to their placement and positioning and why certain materials occur in certain places we know less about. It's probably not centrally related to your talk, but I can't help asking this because of the last videos that you showed. I mean, the behavior one of aggression chasing away another fly and the other one is a different kind of behavior. So is there something like an emotional life of insects? Is there anything known about that? I mean, the term emotion brings up many anthropomorphic. You know, we can't seriously, I can't even for another human being talk about their perceptual sense. You know, it is off to a point where it's almost doesn't become, it's not a science anymore to think in those terms. But in terms of, if we think in terms of complexity or if we say that there are certain basic qualities needed, positive reinforcement, negative reinforcement, whatever is required for learning, give flies any test and they are going to pass it. They have done extraordinary things with flies where you tether them, you can get them to do almost the same thing as what you do in a video arcade where you go and drive a car. You put in a coin, it closes the loop and now you're able to control your visual environment. You can fixate on an object, you can drive towards something. Flies can do that, no problem. And you know, virtual reality arenas in which flies are tethered, you close the feedback loop and flies are able to control their visual environment and they do a fine job at it. You can switch the games, for instance, this is like, you know, you put prism lenses on the eyes and now suddenly everything is backwards and over time humans will adapt to this, flies do too. You can almost throw any test at a fly and they are able to do fine. Now, of course, they're not going to be able to solve equations, but you know, we can't fly. So give or take, but I think anything that has evolved to this time and through such geological upheavals is, you know, I would say smart and has, you know, many of the properties that you might say something like pleasure or pain or, you know, sadness or whatever, as to what it actually is, we can never tell. I can't tell for another human being, so I can't tell for that. But I would say that, yeah, I mean, I'm sure they have some emotional life somewhere by some definition of emotion. Well, mentioning about, I mean, the latter part of your talk, you mentioned about your study on migratory butterflies you were observing. Just curious to know, like, what are the, what is it like that draws them into migration? A lot is known about the migratory birds, but here what is the need? Is it food? Is it the environment? Is it, or just what do you want a different niche for breeding? Well, all of these and temperatures. Temperatures are often the driving factors for what is a good location to breed. And as environments become colder, insects find it harder to survive, so they go to warmer climates. In other places, like in Australia, there's a moth called the Bogong moth, which actually escapes heat and goes to cooler places. So, my temperature drives migration, temperature shifts. Dayland shift drives migration. The requirement to find better food resources is connected to temperature, so that also drives migration. There's many, many things that drive these kind of large scale motion. And in fact, many birds migrate because insects migrate, chasing the food source. But it's an amazing thing. There's a researcher called Charlie Anderson who was working off in Maldives, who has recently made the claim and it's a very, very reasonable claim. It's an extremely lovely paper, which argues that dragonflies are flying from here. In fact, right from outside Bombay, you can see dragonflies going into the sea and going as far as Africa. So, it's extraordinary what they are able to do. And what they are doing is hitching rides on the monsoon wind or the winds, the ITC. With reference to halt years. Uh-huh. Connecting links between butterflies and hot... So, butterflies, lepidopterans in general, are four-winged. They have two pairs of wings. Flies have one pair of wings and one pair of hot tears. Connecting link, I don't know. You know, there are two separate plates in the current point in time. And they must have had a common ancestor at some point, which had four wings. So, two pairs of wings. So, butterflies are specialized. There's another animal of which has hot tears. This is called strepceptor, in which the front wings have become hot tears. And the front wings are doing similar things. I mean, the hot tears are doing similar things there as in the case of flies. There are no further questions. No, there's one. Oh, right, right, right, right, right. Yeah, I... It seems to me that there are different kinds of muscular movement or motions. I mean, there's something that's very little control such as breathing or heartbeat. There's something that's very fine motor control, say speaking, vocalization, and in human speech. So, is there an understanding of whether the different classes and entirely different mechanisms controlling these and the wing movements would come maybe somewhere in between, I'm just curious. I think that's an excellent question. In fact, we're just beginning to get understandings of what constitutes fine control from a motor neuronal viewpoint versus more of this autonomous control, which is like our heart, et cetera. We don't know a whole lot about, for instance, the developmental history of these motor neurons, but a lot is beginning to get described now. There's some really wonderful work coming out of, well, University of Washington was one of the places where a lot of this work was done in the labs of Jim Truman and Darren Williams and so on. And they have been describing these lineages of neurons and trying to track down how these lineages come about and we have a group also working in NCBS on questions of these kind. And maybe it's somewhere in those lineages that you have classes of neurons that are involved in fine control or which innovate, let's say, just the steering muscles and others. I don't know enough about this, but I suspect that those are the kinds of studies that would give us the answers to those kinds of questions. I'm sure there will be more questions, but that can be handled over T, perhaps. So we now break for a short T break and we'll start the next part of the program at 11.35. So please come back to 11.35 and let's thank the speaker once more.