Conus catus, a fish-hunting cone snail (Fig. 1A), delivers venom into its prey by means of a single-use radular tooth (Fig. 1B). The venom is composed of a potent mix of bioactive peptides that, when injected into a fish through the hollow harpoon-shaped tooth, causes tetanus of the body musculature, resulting in a rigid paralysis (1). Although peptide toxins in the venom have been extensively studied (2), the biomechanical mechanisms of tooth insertion and venom ejection have not been determined. Anatomical observations have led to the suggestion that the radular tooth is pushed into prey by muscles surrounding the proboscis lumen (3). In this paper we show that the radular tooth is not pushed directly by the muscles of the proboscis but rather is propelled by a high-speed ballistic mechanism. Small specimens of Conus catus (Hwass, 1792; 3 cm shell length) have a translucent proboscis allowing radular tooth movements to be visualized in situ by using a combination of transmitted-light microscopy with water-immersion optics and high-speed video. Figure 1A illustrates the experimental arrangement. A fish was positioned at the end of a trough where video observations were made. The snail sought the prey by extending its proboscis down a narrow trough in a recording chamber. As the proboscis approaches the fish, hairlike sensory papillae are visible at its tip (Fig. 1C, SP in top panels). Prior to stinging prey, the radular tooth is not held at the end of the proboscis but is positioned with its point 730 ms (about half the length of the tooth) from the end of the proboscis. The tip of the proboscis then contacts the fish, sensing potential prey (Fig. 1C, second panel). The delay between the proboscis first touching the fish and tooth ejection ranged from 240 to 295 ms. With the proboscis held stationary against the fish, the radular tooth is propelled against a constriction of the proboscis lumen (Fig. 1C, arrowheads in the third panels), presumably by pressurization of the fluid space behind the tooth. The slight movements of the radular tooth against the constriction during this “priming” step peak during the 4–5 ms prior to release of the tooth into the fish (Fig. 2). Priming was a consistent feature with a similar time course in 10 feeding sequences captured by high-speed video. During the final millisecond, the radular tooth is explosively propelled into the fish (Fig. 1C, fourth panels). This release step propels the base of the tooth (Fig. 1B, asterisk) to the tip of the proboscis, where it is tightly held by a ring of muscles. The minimum velocity of the tooth during release is approximately 3 ms , but the actual time course of tooth movement is clearly faster than the maximum recording rate employed (1000 frames per second, shutter speed 1/2000). This extremely rapid event exceeds the maximum velocity (2 ms ) for discharge of the cnidarian nematocyst (4). Radular tooth release is one of the fastest known prey capture events and has a time course similar to that of the trap jaw response (0.33 to 1 ms) of the ant Odontomachus (5). Immediately following impalement, the end of the proboscis loses its taper and swells with fluid, especially near the tip where a noticeable bulge appears (Fig. 1C, fourth panels). This fluid, which contains the venom peptides, enters an opening at the base of the tooth and is ejected from both the tip of the tooth and the beginning of the largest radular barb (data not shown and ref. 6, 7). Onset of tetanus in the fish prey is seen within 50 ms of impalement. By gripping the base of the radular tooth, the proboscis is Received 28 April 2004; accepted 19 July 2004. * To whom correspondence should be addressed at Department of Biology M-3, Occidental College, Los Angeles, CA 90041-3314. E-mail: jschulz@oxy.edu Reference: Biol. Bull. 207: 77–79. (October 2004) © 2004 Marine Biological Laboratory