Visualizing vertebrates: new methods in functional morphology

Abstract

Nearly all modern animal functional biology requires motion analysis. As many of the behaviors of greatest interest are very fast (e.g., prey capture and locomotion), image capture has always been of special concern. In the late 19th century, Eadweard Muybridge (and co-workers) solved this problem by rigging a series of still cameras to go off in sequence, ultimately recording the motions of many animals, including humans (Muybridge, 1887). This work provided a treasure trove of functional information, settling, for example, a long-standing dispute (as well as a famous, although possibly mythical, bet made by Leland Stanford) about whether all four feet of a horse leave the ground at any point during a gallop (they do; e.g., Clark, ’31; MacGowan, ’54a). Muybridge’s methods ultimately led to his invention of the ‘‘Zoogyroscope’’ in 1879, the first motion picture projector (later called the ‘‘Zoopraxiscope’’). However, a series of still cameras spread across some distance is impracticable for many animal movements of biological interest; hence, Muybridge’s methods had limited scientific application. About the same time that Muybridge was making his photographs in California, brilliant French physician and scientist, Etienne-Jules Marey (inventor of many biomedical instruments), was perfecting his own system of motion picture capture. Significantly, however, he constructed a single camera that employed a rotating, emulsion-covered disc with an ingeniously synchronized mechanical shutter to capture 12 pictures per second (Marey, 1882a,b). Marey continued to develop various single-camera systems to perfect motion picture taking (or ‘‘chronophotography’’, as he called it) at increasing frame rates, which, unlike Muybridge, he applied specifically to the scientific understanding of animal locomotion (Fig. 1), particularly aerial locomotion in birds (something Muybridge had been unable to do with his multi-camera system; summarized in Marey, 1894, ’02). It is a curious fact that, although Marey is well known to film historians (e.g., Gurnsheim and Gurnsheim, ’55), his remarkable scientific work on animal mechanics is largely overlooked by modern, English-speaking functional morphologists. Visualization and analysis of very rapid movements require a rapid frame rate (i.e., more ‘‘pictures per second’’) and short shutter speeds (the less time the shutter is open, the less motion blur). Marey’s various motion picture cameras were capable of 1/720 sec or shorter shutter speeds—enough to freeze most motions quite well—but his film disc system limited the number of frames that could be taken (12–24) and therefore, the duration of a behavioral sequence that could be filmed. With the Eastman Company’s invention of nitrocellulose film in 1889 (McGowan, ’54b), the way was paved for faster frame rates and longer run times. In 1891, the Edison Company exploited this advance with their introduction of a new motion picture camera (the ‘‘Kinetograph’’) and a projector to display the films to audiences (the ‘‘Kinetoscope’’). It is this basic form of technology that led to modern cinematography (Musser, ’95) and ultimately, the scientific application of high-speed film to analysis of animal kinematics. However, film systems are cumbersome and have many limitations of speed, lighting, record time and resolution (and worse—one has to wait for the film to be developed before seeing the results!). During the last 25 years, film systems have been replaced by video systems (first analogue, then digital), and recently, these have achieved astounding frame rates and levels of resolution, making visualization of even the most fleet and fleeting of animal movements possible. As prices have come down, high-speed video systems have become standard equipment in most laboratories that study the dynamics of animal movement. The story might end there except for one unfortunate fact— most animals are opaque. By definition, anatomists are primarily