Sherborn's Index Animalium: New names, systematic errors and availability of names in the light of modern nomenclature

Abstract

This study is aimed to shed light on the reliability of Sherborn's Index Animalium in terms of modern usage. The AnimalBase project spent several years' worth of teamwork dedicated to extracting new names from original sources in the period ranging from 1757 to the mid-1790s. This allowed us to closely analyse Sherborn's work and verify the completeness and correctness of his record. We found the reliability of Sherborn's resource generally very high, but in some special situations the reliability was reduced due to systematic errors or incompleteness in source material. Index Animalium is commonly used by taxonomists today who rely strongly on Sherborn's record; our study is directed most pointedly at those users. We recommend paying special attention to the situations where we found that Sherborn's data should be read with caution. In addition to some categories of systematic errors and mistakes that were Sherborn's own responsibility, readers should also take into account that nomenclatural rules have been changed or refined in the past 100 years, and that Sherborn's resource could eventually present outdated information. One of our main conclusions is that error rates in nomenclatoral compilations tend to be lower if one single and highly experienced person such as Sherborn carries out the work, than if a team is trying to do the task. Based on our experience with extracting names from original sources we came to the conclusion that error rates in such a manual work on names in a list are difficult to reduce below 2-4%. We suggest this is a natural limit and a point of diminishing returns for projects of this nature.

Keywords: Sherborn.

Figure 1.

Number of names. Total number of taxonomic names listed in Sherborn’s Index Animalium and in AnimalBase (2011) plotted in 5-year intervals. It must be taken into account that Sherborn’s numbers included a proportion of 30% of names that were not new, while in AnimalBase this proportion was much lower (less than 5 %).

Figure 2.

Languages used in early zoological literature. Language analysis of 2100 arbitrarily selected binominal zoological works published between 1758 and 1850.

Figure 3.

Example of a new name established in a chaotically arranged early zoological work. Detail of Hartmann (1821: p. 231), with the original description of Helix ruderata Hartmann, 1821 (currently Discus ruderatus). For the non-insider it is very difficult to see that a new name was established here. It is necessary to understand precisely the content of the German text: “s. meine Tab. II. f. 11. Der vorigen ähnlich, aber aufgeblasener, rauher, weniger Umgänge“ (= see my plate II, figure 11. Similar to the previous one, but more inflated, rougher, fewer whorls).