Psammornis rothschildi ANDREWS , 1912

Psammornis rothschildi ANDREWS, 1912

Text-figs 3–6, 8–11

S p e c i e s d i a g n o s i s. Based on description by

Andrews (1912): Egg shells 3.2–3.4 mm thick.

L e c t o t y p e. (designated herein) NHMUK E/1963.9.218a eggshell fragment and thin sections (NHMUK PV A 1334 a and b) cut from same (mentioned as “the type shell” by Sauer 1969) ( Text-figs 3, 6a, b, 8). Sauer (1969) mentioned two specimens as “types” but there is no guarantee that the two fragments came from the same egg, clutch, individual bird to even the same species. To avoid possible confusion, we nominate one of the specimens as lectotype, and refer the other specimen to the species.

Te n t a t i v e l y r e f e r r e d s p e c i m e n s. 1) NHMUK E/1963.9.218b fragment of eggshell and thin sections cut from it (NHMUK PV A 1334 c and d), from near Touggourt, Algeria ( Text-figs 4, 6c, d, 9). 2) NHMUK E/1963.9.218c ( Text-figs 5, 10) fragment of eggshell reportedly collected near Touggourt, Algeria ( Rothschild and Hartert 1912) but possibly from Temassinine, Algeria.

L e c t o t y p e l o c a l i t y. 20 miles east of Touggourt towards El Oued, Algeria ( Rothschild and Hartert 1912).

A g e. Imprecise, probably latest Miocene or Plio- Pleistocene (see Buffetaut 2022, for discussion on the age of the specimens attributed to Psammornis ); not Eocene as published by Rothschild (1912).

E t y m o l o g y. Psammornis – from Greek Ψάμμος

(Psammos) = Sand + ὄρνις (ornis) = bird.

D a t e o f p u b l i c a t i o n. With one exception, all scientists who have writtten about Psammornis have cited Andrews as the author and 1911 as the date of publication. The cover page of the congress proceedings (Verhandlungen) gives the date of publication as 1911, but, as was pointed out by Hartert (1927) it was not issued until the beginning of 1912, which, for the establishment of “priority” of zoological names (the case with Psammornis ) must be the date of publication. Hartert (1927) concluded that the date of creation of the nomen Psammornis rothschildi by Andrews was 1912, because, as he wrote, “The volume, though dated 1911, did not appear before January or February 1912 ”.

The preprint (Sonderabdruck but called Bericht in the literature and on museum labels) of Rothschild’s paper (with Andrews’ paper as its appendix) was reviewed in the January 1912 issue of the Ibis (pp. 202–203). These preprints are usually dated from 1911 (but from 1910 in the Ibis). It is thus possible that Psammornis rothschildi dates from the Sonderabdruck of 1911 (?1910), not from the Verhandlungen of 1912. However, in view of Hartert’s (1927) statement, we accept the Verhandlungen as the version that established the name of the taxon.

The nomen Psammornis rothschildi was “pre-published” by Rothschild and Hartert (1912; the cover page of the issue gives the date as 1911–1912, but the issue was released in 1912). Having evidently seen a preprint or proofs of the Andrews’s paper, the authors (incorrectly) cited its creation as being in the Bericht über den V. Intern . Ornith. Kongress, pp. 150 and 169–73 (Bericht = Report, whereas the Proceedings appeared as Verhandlungen). The incorrect “Bericht” citation was repeated by Bédé (1919) and is written on the labels of the thin section slides ( Text-fig. 6).

Psammornis rothschildi , as cited by Rothschild and Hartert (1912) is a nomen nudum because it was not accompanied by any description or designation of type material. Furthermore, the citation of the name (variously as Psammornis or Psammornis rothschildi ) by Rothschild (1912) on pages 144, 146, 147, 150 and 167 of the congress proceedings (i.e., in the article in the Verhandlungen) does not give the author page priority because all these mentions are nomina nuda. The valid erection of the genus and species was by Andrews 1912, on page 173 of his article (pp. 169– 174) in the proceedings.

D o c u m e n t a t i o n a n d c u r a t i o n. The original hypodigm of Psammornis rothschildi consisted of two fragments of eggshell ( Andrews 1912) even though Rothschild and Hartert (1912) reported that there were three fragments from the site. Andrews (1912) also mentioned the presence of thinner eggshells from the same location attributable to the extant ostrich, Struthio View in CoL , but no details of their morphology were provided.

At the Natural History Museum, Tring, three eggshell fragments labelled “Type material” Psammornis rothschildi ? are curated under the register No. NHMUK E/1963.9.218 (i.e., entered in the register 54 years after the year of collection). The main label ( Text-fig. 2) refers to the fact that Andrews’ “type” was TWO fragments but that the sample consists, in fact, of three fragments. The thin sections made from two of the fragments that were previously kept in London, have the register No. NHMUK PV A 1334 ( Sauer 1969). Note that the main label in the Tring Museum articulates doubts about the precise nature of the type material, and it is for this reason, among others, that it is necessary to nominate a lectotype for the species. Small red cardboard labels with the fossils give the date of publication of the genus and species as 1912 ( Text-fig. 2) .

Inspection of the table at the end of Schönwetter’s (1929) paper shows that he listed two shards of “ Psammornis rothschildi Andr. ” from Touggourt (3.40 and 3.30 mm thick, that he specified as typus), and several shards (mehrere Scherben) from South of Biskra, 3.20 mm thick (labelled as having been collected by Hilgert). The locality data of the latter specimen(s) is probably erroneous, as explained below .

In a separate entry in the same table published by Schönwetter (1929), is the mention of several shards of eggshells from South of Biskra that measure 2.50 mm thick listed as “ Struthio spec. extinct?”. Our own measurements of the thickness of 41 fossil eggshell fragments from South of Biskra curated at Tring ( NHMUK E/1963.9.236), give a range of variation of 1.8 to 2.6 mm, with the commonest measurement (15 fragments) at 2.5 mm. It appears that, at the time of Schönwetter’s visit to Tring , there may have been a mixup such that the supposed “third” specimen from Touggourt had became mixed in with the thinner shells from Biskra, and that it was subsequently re-associated with the other two fragments from Touggourt. The preservation characters of the specimen ( Text-fig. 5) differ from the fossils from Biskra, which are darker chocolate coloured and show a greater degree of erosion of the shell surfaces. However , the fact that the third specimen labelled Touggourt has some aeolianite adhering to it suggests that it may have been collected at Temassinine, Algeria (see taphonomy section below) .

Ta p h o n o m y. Andrews (1912) and subsequent authors

( Schönwetter 1929, Sauer 1969) reported that the inner and outer surfaces of the eggshells of Psammornis rothschildi had been abraded by sand blasting, thereby reducing their thickness, but our own examination of the surfaces of the shells suggests that such is not the case, the polishing and abrasion is weak and has not greatly affected the thickness of the shells ( Text-figs 8–10).

The two specimens from Touggourt studied by Andrews (1912) are free of matrix, but the third specimen labelled as coming from Touggourt, retains small patches of indurated, slightly reddened aeolianite. The aeolianite adhering to this specimen resembles that which occurs on several of the specimens from Temassinine (Fort Flatters) southern Algeria, collected by Fromholz. In addition, the shell is lighter in colour than the other two specimens from Touggourt. These observations raise the possibility that the “third” Touggourt specimen came from a different locality, possibly Temassinine. Hartert (1927) and Schönwetter (1929) briefly mentioned the material from the site (as Temassin (sic) by the latter author). The 12 specimens from Temassinine curated at Tring (NHMUK E/1963.9.221) range in thickness from 2.5 to 3.0 mm, and the third specimen supposedly from Touggourt is 3.2 to 3.3 mm thick, slightly greater than the thickest shell from Temassinine. Given the uncertainty, we prefer to refer the third specimen labelled as coming from Touggourt to Psammornis rothschildi , but not to include it in the type series.

Sauer (1969) estimated that the original thickness of the Touggourt shells could have been about 4 mm, which, if correct would imply that about 0.7–0.8 mm of matter has been removed from their surfaces. However, the lectotype shows well preserved pore slits and the pores in their depths are clearly visible over much of the shell ( Text-fig. 8). The paralectotype shows some effects of abrasion, including superficial scratches, and some enlargement of the pore slits and depressions, possibly due to repeated exposure to dew ( Text-fig. 9), but the damage has not greatly altered its thickness. The third specimen labelled as coming from Touggourt is well preserved and even retains small patches of indurated aeolianite on its external surface ( Text-fig. 10). The pore slits are moderately well preserved, as are the pores in some of the depressions. Our measurements of the thicknesses taken at three to four places on each of the three eggshells labelled Touggourt, are 3.2 and 3.3 mm.

M o r p h o l o g y. The external surfaces of the eggshell fragments from Touggourt show irregularly spaced straight to slightly angular, quite short, slit-like depressions arranged with their long axes sub-parallel to each other or in a meandering pattern. The slits or depressions usually contain two or three pores, sometimes as many as four or five. Some of the depressions have been modified post-mortem by erosion, but the alteration does not affect the entire surface. However, care needs to be taken to avoid interpereting some depressions as pore complexes when that might not be the case.

The surface microstructure of the eggshells of Psammornis rothschildi was described in detail, but not illustrated, by Andrews (1912). In the thin sections illustrated by Sauer (1969) the more structured inner and outer layers of the eggshells are observed to be separated from each other by a layer of finer-grained amorphous material. The amorphous layer corresponds to what was called the “spongy” layer by Andrews (1912), Sauer (1969) and Louchart et al. (2022). The outer layer is about twice as thick as the amorphous layer, which is slightly thinner than the innermost layer ( Text-fig. 11). The relatively great thickness of the spongy layer in Psammornis eggs represents a major difference from eggshells of Struthio (sensu stricto).

B i o s t r a t i g r a p h y. When collected, most of the North African and Arabian fossils hitherto attributed to Psammornis rothschildi were not found in situ in datable deposits ( Andrews 1912, Rothschild and Hartert 1912, Hartert 1927, Lowe 1933a). Because of this, age estimates have varied widely, ranging from Eocene to Pleistocene ( Andrews 1912, Rothschild 1912). An exception is the material from Chebket Safra, Tunisia ( Choumowitch 1951, Buffetaut 2022) which is likely to be of Messinian age, although previously the deposits (the Segui Formation) from which the eggshells were collected were originally correlated to the Pontian (in the old sense of the term as the continental equivalent of the Pliocene, but in reality of Late Miocene age). Choumowitch (1951) reported that the eggshell fragments from Chebket Safra are regularly 3 mm thick (“Ces fragments ont très régulièrement trois millimètres d’épaisseur, c’est à dire un millimètre du plus que l’oeuf d’autruche courant”).

Pickford et al. (2023) described eggshells (2.0– 2.7 mm thick) excavated from Bou Hanifia, Algeria, from deposits that were for a long time correlated to the Pontian ( Arambourg 1959), but which are more likely to be of Ventian age ( Pickford and Chaïd-Saoudi 2024). The Bou Hanifia specimens were initially attributed to Diamantornis laini , but following detailed examination of the Omani collection, it is now considered to be more likely that they belong to a small species of Psammornis . A similar reattribution of the eggshells (2.0–3.0 mm thick) from the Marsawdad Formation, Rub al-Khali, Oman, is likely.

Even though there remains a wide margin of error in the age determinations of eggshells attributed to Psammornis , the specimens from Chebket Safra, Tunisia, suggest that Psammornis rothschildi is of latest Miocene (Ventian) age, while the slightly thinner specimens from Bou Hanifia, south of Biskra, and Temassinine (ex Fort Flatters) ( Hartert 1927, Schönwetter 1929) in Algeria, and the Marsawdad Formation in the Rub al-Khali, Oman ( Pickford et al. 2023) are likely to correlate to the Ventian or Zanclean.

Thus, even though some doubt remains about its stratigraphic range, it is considered that the type material of Psammornis rothschildi is likely to be of Ventian age.