From Collector to Collection: What Information can be Gleaned from Historic Museum Collections?

Written by Sarah King, Curator of Natural Sciences, York Museums Trust

Geology collections by their very nature are composed of historically acquired specimens. Each specimen will have gone through a process before being curated as part of a museum collection, and, especially for the material acquired decades or hundreds of years ago, it is often impossible to unpick this history. All too often, much information is missing, and there is no way to know the reasons for, or rationale behind, the specimens that make it into a collection. Individual specimens can be used in their own right for a multitude of things, even with sparse data, but what if the opposite is the case? If a large number of specimens are donated together, can they be reliably used collectively to offer more information? And are there any lessons for curators acquiring material? I present here a brief case study, following the process of studying a collection of plant fossils in a museum collection, but with the added bonus of full, original records of specimen collection and interpretation.


David Davies (1871-1931) was a colliery manager in south Wales during the early 20th century. Davies had a great interest in the plant fossils found by his miners in the roof shales above the coal seams being worked, and had every one brought up for him to identify, with the assistance of his great friend, the well-respected British Geological Survey palaeobotanist Robert Kidston (Thomas, 1986). He published a comprehensive paper on the eastern sections of his area (‘Glamorganshire’) in 1929, with full records of over 200,000 specimens he had found, and an impressive early assessment of the climate and various environments occupied by these cyclic, swampy ecosystems (see Figure 1).

Figure 1: An interpretation of a typical late Carboniferous swamp, made around the same time as Davies was working. Image in the public domain.

As part of his life’s work, he had a proportion of his finds sent to the Amgueddfa Cymru – National Museum Wales (NMW) in Cardiff (see Figure 2), with the intention of their being used as a teaching collection (North, 1935).

Figure 2: Some examples of the David Davies plant fossil specimens held at National Museum Wales, Cardiff. a) Cyathocarpus aff. arborescens (Brongniart) Weiss – a fern (in group Filicales). b) Neuropteris ovata Hoffmann – a pteridosperm, or seed fern. c) Lepidophloios laricinus Sternberg – a lycophyte, or lycopsid. © Collections Services, National Museum Wales, used with permission.

Sadly, Davies died in 1931, before he could complete a similar treatment of the western sections. Today, a subset of around 16,000 specimens from his huge collection resides at NMW, along with his notebooks. The systematic and thorough work of Davies in identifying his finds to species level consistently, regardless of quality or size, and recording their immense number in a published paper, alongside the availability of the original specimens for study, offers a rare opportunity to investigate the suitability of historic museum collections with regard to analyses such as relative abundances, ecological composition and turnover.


Davies recorded fossils based around 29 coal seams, spanning roughly 312-306 million years ago in the Pennsylvanian (later Carboniferous) sub-period. He numbered the seams 1-29, with 1 being the youngest (highest). For this study I extracted the data for the uppermost seam, the Llantwit no. 1. Data were taken from Davies’ paper (over 11,000 specimens), and the museum specimens (numbering 634 – around 6% of the original number), and divided into related plant groups.  In order to do this, Davies’ taxonomic nomenclature was examined. It was found to be remarkably robust, and required little updating to bring it up to modern standards. Additionally, the species list was pared down to just include the most diagnostically useful plant organs (after Cleal, 2007; Thomas, 2007): fossilised plants are nearly always completely disarticulated, leading to different names being applied to organs from the same original whole plant (Bateman and Hilton, 2009). Therefore, to make meaningful, biological analyses of the preserved remains of the community, avoiding over-representation of each component, it is necessary to attempt to compare like with like, by considering equivalent organs from each species. In most cases this proved to be foliage, as with fossil plants the taxonomy here is far more developed and comparable. An exception to this is the stem morphology of the lycopsids (which include Lepidodendron, the ‘scale tree’), which is used in place of foliage because with this group, the leaves are indistinguishable between species.


Figure 3: a) Davies’ published data on the Llantwit No. 1 seam, comprised of over 11,000 specimens. Sphenopteris at 0.4% of the total, as documented by Davies, could be either filicales (fern) or pteridosperm (seed fern): unfortunately it is impossible to tell in the paper. b) Data compiled from the collections at the NMW from the Llantwit No. 1 seam, comprising 634 specimens.

Results and Discussion/Conclusions

The order of relative abundances of plant groups is not vastly different between Davies’ published data and the subset of material at the NMW (see Figure 3): pteridosperms (extinct seed ferns), lycophytes (including Lepidodendron) and sphenophytes (including the horsetails) dominate, and sphenophylls (an extinct group related to horsetails) are the smallest component. Pteridosperms at 62.8% (from the collections) and 61.8% (from the paper) are the most similar in relative abundance, but after this point the percentages of the respective plant groups vary widely. The totals from the NMW show evidence of bias: an over-representation of the more minor plant groups seen in the paper, probably a conscious effort to retain a viable collection of each of the groups regardless of true abundances.

While this means that each species originally found in association with the seam should definitely be represented in the museum collection, with enough specimens to allow a degree of systematic work, it does confirm that, without explicit information to the contrary, museum collections cannot be used for accurate species abundance studies: only presence/absence can reliably be gleaned. Much more analysis and comment is possible also – see King et al 2011.

It is of course only possible to carry out this sort of analysis due to the incredible diligence of Davies and his workforce in collecting and documenting all the fossils they encountered. There is always a story behind the specimens that make it into a permanent museum collection – and those that are missing. The Davies collection is a wonderful legacy that reinforces the commonly uttered refrain of the curator when presented with a potential acquisition: ‘What data does it have with it?’. As the Davies collection shows, there can never be too much data associated with a specimen. When acquiring newly-collected specimens, time should always be taken to record any and all possible information about them, and anything discarded. Think about why the material was collected – is it the largest, first or best? What has been left behind, and how do you know? How much can an individual specimen tell you about the species, area or collecting agenda, and can you find out more about its context? Perhaps researchers will be working on your material a century from now, thanking you for doing just that.

Bateman, R.M., Hilton, J. 2009. Palaeobotanical systematics for the phylogenetic age: applying organ-species, form-species, phylogenetic species concepts in a framework of reconstructed fossil, extant whole-plants. Taxon, 58, 1254-1280.

Cleal, C.J. 2007.  The Westphalian-Stephanian macrofloral record from the South Wales Coalfield.  Geological Magazine, 144, 465-486.

Davies, D. 1929.  Correlation, palaeontology of the Coal Measures in east Glamorganshire.  Philosophical Transactions of the Royal Society of London, Series B, 217, 91-153.

King, S.C., Cleal, C.J., Hilton, J. 2011. Common ground between two British Pennsylvanian wetland floras: Using large, first-hand datasets to assess utility of historical museum collections. Palaeogeography, Palaeoclimatology, Palaeoecology, 308, 405-417.

North, F.J. 1935.  The fossils and geological history of the South Wales Coal Measures: Some aspects of the work of the late David Davies.  Proceedings of the South Wales Institute of Engineers, 51, 271-300.

Thomas, B.A. 1986.  In search of fossil plants: the life and works of David Davies (Gilfach Goch). National Museum of Wales, Cardiff, 54 pp.

Thomas, B.A. 2007. Phytogeography of Asturian (Westphalian D) lycophytes throughout the Euramerican belt of coalfields. Geological Magazine, 144, 457-463.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s