A spectacular, well-observed Leonid meteor of visual magnitude −14.3 appeared on 17 November 1998 and left a lingering trail, dubbed the Glowworm, that was well studied. From a location on Kirtland Air Force Base, near Albuquerque, New Mexico, we obtained CCD images of the trail from 94 to 203 s after the meteor and recorded a video with an intensified camera for even longer. From information obtained with a sodium lidar half an hour after the meteor, we have determined that a gravity wave with a vertical wavelength of 2.4 km was responsible for the right-angled appearance of the trail. The trail ended abruptly at 85 km, and its uppermost altitude may have been no greater than 91 km. We designate the Glowworm a Type I trail: one that is wide (1 km), cloudy in appearance, has high diffusion rates (800 m² s⁻¹), high total line emission rates (1.5×10¹⁸ photons m⁻¹ s⁻¹), and is optically thicker than Type II trails. The lower parts of the Diamond Ring, another Leonid lingering trail that appeared 38 min earlier than the Glowworm, define the Type II trails, which appear as narrow, optically thinner parallel trails, with low diffusion rates (12 m² s⁻¹) and total line emission rates (1–3×10¹⁶ photons m⁻¹ s⁻¹). No explanation is offered for the two orders of magnitude difference in these quantities. The Glowworm meteor produced infrasound [ ReVelle and Whitaker, 1999 ], from which a meteoroid mass estimate of 522 g was made. We compare our photometry to a detailed numerical modeling of the shape of the trail and emission from the Glowworm made by Zinn et al. [1999] , who find that the largest contributors to emission recorded by our CCD and video cameras are atmospheric O₂ vibrational bands. Compared to our measurements, their calculated emission is too high by two orders of magnitude, but since most of O₂ emission may be absorbed by atmospheric O₂ before it reaches the ground, this may indeed be the primary contributor to the observed flux. Although the calculations of Zinn et al. lead to a hollow cylinder appearance which may be appropriate for the Glowworm, it is not pronounced enough to account for the complete darkness between the parallel structures seen in Type II trails. An upper limit to backscattering from dust of 3.7 × 10⁻⁵ of the expected return was found from directing a 180 W copper vapor laser at the Glowworm.