Detection of lost calamus challenges identity of isolated Archaeopteryx feather

In 1862, a fossil feather from the Solnhofen quarries was described as the holotype of the iconic Archaeopteryx lithographica. The isolated feather’s identification has been problematic, and the fossil was considered either a primary, secondary or, most recently, a primary covert. The specimen is surrounded by the ‘mystery of the missing quill’. The calamus described in the original paper is unseen today, even under x-ray fluorescence and UV imaging, challenging its original existence. We answer this question using Laser-Stimulated Fluorescence (LSF) through the recovery of the geochemical halo from the original calamus matching the published description. Our study therefore shows that new techniques applied to well-studied iconic fossils can still provide valuable insights. The morphology of the complete feather excludes it as a primary, secondary or tail feather of Archaeopteryx. However, it could be a covert or a contour feather, especially since the latter are not well known in Archaeopteryx. The possibility remains that it stems from a different feathered dinosaur that lived in the Solnhofen Archipelago. The most recent analysis of the isolated feather considers it to be a primary covert. If this is the case, it lacks a distinct s-shaped centerline found in modern primary coverts that appears to be documented here for the first time.

The length and width of the calamus halo matches that of the original published description 2 (Fig. 1). Microscopic examination revealed past preparation had engraved around the outline of the feather and inadvertently prepared away the calamus at some unknown point in the past. Thus, the recovered geochemical halo is a chemical breakdown residue fluorescing immediately beneath the surface of the original carbon or manganese dioxide film. The feathers are clearly defined in many Archaeopteryx skeletons 1 . The feather impressions from some of the more complete specimens allows for detailed morphologic measurements 1,18 . The general morphology of Archaeopteryx feathers is considered similar to modern birds, allowing cautious comparisons with living taxa 1 .
As in extant birds, the primaries of Archaeopteryx are characteristically straight and have vane asymmetry 19 . Their straightness does not match the isolated feather and they are also generally more asymmetrically vaned. The isolated feather's identification as a primary feather has also been historically argued against 1,7 . Archaeopteryx lacks a bastard wing (alula) 1 , so the identification of the isolated feather as an alula feather of Archaeopteryx can be excluded.
The isolated feather is also not a tail feather (rectrix) of Archaeopteryx. The distal rectrices of Archaeopteryx are extremely long and symmetrical in outline at the tip (eleventh specimen: Fig. 2E of 18 ), two features absent in the isolated feather. The isolated feather shares a general asymmetry in outline and rachis position with the lateral rectrices, but the curvature of the rachis is too severe in the isolated feather to form the frond pattern seen in Archaeopteryx (eleventh specimen: Fig. 2F of 18 ). The tail feathers of the London specimen lack asymmetrical vanes, which also contrasts with the morphology of the isolated feather 1 .
The secondary feathers in the known Archaeopteryx specimens are the closest matches to the general feather outline of the isolated feather. Unfortunately, no other feathers stand alone in other Archaeopteryx specimens with feather preservation, but measurements of the isolated feather can be compared to the secondaries of the Berlin specimen, which preserves the most complete wing feathering of Archaeopteryx 1,14 . The outline of the isolated feather was superimposed onto a version scaled to match the width of the most similar secondary feather in the Berlin specimen (Fig. 2). This comparison reveals that the isolated feather is 1/3 shorter than required to scale to the secondaries of the Berlin Archaeopteryx wing. Unfortunately, the specimens larger than the Berlin specimen (London and Solnhofen) as well as the smallest urvogel (Eichstätt) both have poorly preserved feathering 1 , so this cannot be compared across ontogeny.
A range of secondary feather counts has previously been reconstructed along the ulna of the Berlin specimen (ten 20 , twelve 21 , fourteen ( Fig. 6.18 of 1 ) and twelve to fifteen 22 ), but the reliability of these counts has been questioned 18 . Scaling the isolated feather to match the length and spatial overlap in the wing of the Berlin specimen (Fig. S7) shows that 7 secondaries could fit along the wing, significantly fewer feathers than past reconstructions. If the isolated feather was from a subadult as suggested by Wellnhofer 1 , then the feather count on the shorter ulna would be even less. As mentioned, this cannot be compared across ontogeny as the largest and smallest Archaeopteryx specimens (Solnhofen and Eichstätt) have poorly preserved feathering 1 . Nevertheless, these data raise questions about the fit of the isolated feather to the wing of Archaeopteryx. The remaining possibilities for the isolated feather are as a covert or a contour feather. However, a determination is less straightforward. Little is known about the contour feathers of Archaeopteryx, although modern contour feathers typically have less robust calami than the isolated feather. As a covert, the isolated feather is very different to those of extant birds. In living birds, the secondary coverts attach to the calamus of the secondary flight feathers at an angle (Fig. S8). This configuration necessitates a shorter calamus than the primary coverts, which are in place alongside the primary feather calamus. The robust calamus of the isolated feather is therefore too large for a secondary covert, so this identification is not supported. The most recent analysis of the isolated feather considered it to be a primary covert 8 . The size-normalized calamus-rachis centerlines of primary coverts from 24 modern birds, including those of different body sizes, were compared to the isolated feather (Fig. 3). All possess a calamus-rachis centerline that curves towards the leading edge of the wing from the centerline of the calamus, unlike the rachis centerlines of the other feather types present in the same wing specimens 7,19,23,24 (Figs 3,  S3-S6). This 'S-shaped' centerline described here for the first time, appears to be a defining characteristic of primary coverts across a very broad range of modern species, including the palaeognath tinamou. In contrast, the centerline of the isolated Solnhofen feather curves strongly toward the wing's trailing edge (see blue line in Fig. 3) so does not match the morphology of primary coverts in modern birds 7,19,23,24 .
In summary, the isolated feather is not conformal to known Archaeopteryx specimens as a primary, secondary or tail feather. Its preservation as a dark film also differentiates it from all other known specimens 1,6 . The isolated feather as argued here lacks any close morphological connection to the 11 or 12 known Archaeopteryx skeletons   Fig. S3) whose wing has been cited as the isolated feather's closest modern match 1,7 . In brown is the centerline trace from a modern Undulated Tinamou (Crypturellus undulatus UWBM 71526, Fig. S4), which belongs to the only groups of extant palaeognaths with flight capabilities. The yellow zone represents the area covered by the traces of all 24 measured feathers, including a 1.5% error zone allowing for taphonomic flex (see Fig. S1). In all cases the isolated feathers centerline is a large departure from modern primary coverts. (see status of Haarlem specimen 4,5 ), but not all feathers of Archaeopteryx are known. However, based on known feather preservation in Archaeopteryx, this study raises the possibility that the isolated feather may belong to another basal avialan or even a non-avialan pennaraptoran, increasing the low theropod diversity of the Solnhofen Archipelago 1,4,[25][26][27] . This hypothesis would be in agreement with comments made in Opinion 2283 (Case 3390) of the ICZN Commission 3 as well as the recent removal of the Haarlem specimen from Archaeopteryx 4 . The feather remains an enigma so we caution against the isolated feather's association with Archaeopteryx.

Material and Methods
Archaeopteryx specimens studied.
Scientific RepoRts | (2019) 9:1182 | https://doi.org/10.1038/s41598-018-37343-7 Primary covert feather analysis was performed from photographs. These were sourced from museum collections and the Vogel Federn and Michel Klemann online feather atlases 23,24 . The feathers of the latter two collections were flat bed scanned (see Supplementary Materials for discussion of flattening-related feather taphonomy). Feather centerlines were overlaid in Photoshop CS6, all centerlines were scaled to the same length.