detailed drawing of specimens. The basic item of machinery, the camera lucida, is no different now from the model that Walcott used, and not much improved from its original invention by the mineralogist W. H. Wollaston in 1807. A camera lucida is, basically, a set of mirrors that can focus the image of an object onto a flat surface. You can attach a camera lucida to a microscope and cast the image under the lens onto a piece of paper. By simultaneously viewing the specimen and its reflection on paper, you can draw the animal without moving your head from the eyepiece. Whittington and his team adopted the procedure of drawing every specimen, at very large scale, for any species under investigation. You can study a set of drawings together, but you cannot easily make simultaneous observations on numerous tiny specimens, all needing magnification.
Whittington applied his camera lucida and skill in drafting to a set of methods all linked to his central recognition that the Burgess fossils retained some three-dimensional structure, and were not just flattened sheets on bedding planes. I shall illustrate the power of these simple procedures by showing their usefulness in the study of the largest Burgess arthropod, the species that Walcott named Sidneyia inexpectans to honor his son, who had found the first specimen. (I choose Sidneyia because David Bruton’s 1981 monograph on this genus is, in my opinion, the most technically elegant and attractive publication of the entire series by Whittington and his associates.) Consider the three main operations:
3.1. An attractive plate of Burgess photographs from Walcott’s 1912 monograph on arthropods. The photographs are extensively retouched. Canadaspis is at top left; Leanchoilia at bottom.
3.2. The best unretouched photo ever taken of a Burgess Shale organism. Des Collins took this photograph of a Naraoia , preserved in side view. This specimen does not come from Walcott’s quarry, but from one of the dozen additional localities for soft-bodied fossils recently found by Collins in the same area. Specimens from Walcott’s quarry do not photograph this well.
1. Excavation and dissection . If Walcott had been right, all anatomy would be compressed into a single film, and the task of reconstruction would be akin to reviving a cartoon character squashed flat by a steam-roller. But what works for Sylvester the cat in a world of fantasy cannot be duplicated for a slab of shale.
Fortunately, the Burgess fossils do not usually lie on a single bedding plane. Engulfed by the mud that buried them, the animals settled into their tombs at various orientations. The mud often infiltrated and sorted their parts into different microlayers, separated by thin veils of sediment—carapace above gills, and gills above legs—thus preserving some three-dimensional structure even when the muds became compressed later on.
By using small chisels or a very fine vibro-drill, not much different from the model in your dentist’s office, upper layers can be carefully removed to reveal internal parts beneath. (As these layers are often but microns thick, this delicate work can also be done by hand and with needles, grain by grain or flake by flake.)
3.3. Reconstruction of Sidneyia from a three-dimensional model built in sections by Bruton. (A) The entire animal. (B) The model in six segments, starting from bottom left—the head with its ventral covering plate below, the body in three sections, and the tail piece. (C) The head and front part of the body connected, with the head in the background and to the right. Note the biramous appendages with their walking legs below and gill branches above.
Some arthropods are fairly flat, but Sidneyia , as the reconstruction shows (figure 3.3), possessed considerable relief; its carapace, or outer covering, formed an arched semicylinder over the soft parts beneath. * In some specimens the underlying gills and legs protrude through a broken carapace, for natural
Chip Hughes
Brian Moore
Neeraj Chand
Kam McKellar
Marion G. Harmon
John le Carré
A. L. Summers
Antal Szerb
Tim Tharp
Flying Blind (v5.0)