On Horses, In Solidarity

Joyce C. Havstad writes…

I cannot speak to all or even most of what was tragic about the burning of Brazil’s National Museum in September. But it saddened me greatly.  In an effort to express my sympathy and solidarity, today’s post will discuss two salient points: the value of collections and the difficulty of maintaining them.  Perhaps unexpectedly, I’m going to open my discussion with talk of equine evolution—and fans of Stephen Jay Gould (1980, 1987, 1996) will likely be familiar with how this horse’s tale begins.

In 1859, Charles Darwin published On the Origin of Species by Means of Natural Selection, or the Preservation of Favoured Races in the Struggle for Life—and many scientists were only partially convinced by the monograph, to varying degrees.  Many accepted the truth of evolution, but denied that natural selection was the primary agent of such biological change.  The Russian paleontologist Vladimir Onufrievich Kovalevsky (1842–1883) was the first to convincingly document Darwinian natural selection acting on fossil lineages in a manner conducive to proper evolutionary transformation.

In a series of works (published 1873–1877), Kovalevsky used fossil horse specimens from throughout Europe in order to meticulously link changes in fossil organism morphology to changes in external environment—in other words, he offered paleontological proof of adaptation.  On Kovalevsky’s account of horse evolution, a small, many-toed, leaf-browsing woodland ancestor from the Eocene (Paleotherium) was driven by the evolution of grasses and grassland towards a three-toed early Miocene form (Anchitherium) through a late Miocene version (Hipparion) and, eventually, horses evolved into the large, single-toed, hard-toothed grazers and gallopers of modernity (Equus).

Thomas Henry Huxley came to the same conclusion about horse evolution that Kovalevsky did, perhaps even a little earlier, but Huxley gave credit for the case to the Russian, due to Kovalevsky’s superior documentation of the transformation.  And this is when we glimpse the first twist in the tale: despite converging on the same account at around the same time, both scientists were—in one very important sense—utterly wrong about what the succession of European specimens showed.  Huxley discovered the error rather abruptly, during his fall 1876 tour of America cities, universities, and fossil collections.  He arrived in the States on August 5 and, by the second half of September, Huxley was telling a revised story of horse evolution to crowds in New York (see Jensen 1988).

It turns out that Kovalevsky was correct—paleontologists still agree—that horse evolution was driven by the evolution of grasses and grassland, and that this succession of evolutionary changes is reflected in the collection of European horse fossil specimens.  But the specimens in that collection are not themselves linked by direct ancestor-descendant relations.  Much of early horse evolution happened in the Americas, not in Europe—so that is where the successively changing specimens are linked by direct ancestor-descendant relations (initial documentation in Marsh 1874).  Migratory offshoots from the evolving American populations repeatedly ended up in Europe, only to die out there.  That is until approximately 10,000 years ago, at which point the tables turned: New World equine populations disappeared, and horse evolution proceeded via domestication of stock from the rather limited equine populations of the Old World, instead.  (See MacFadden 2005 for more.)

Gould enjoyed using this story—a literal “textbook case” of evolution occurring in a fossil lineage, perhaps the most famous example of such—in order to debunk popular but naïve ideas of evolution as linear, tidy, and progressive.  In now-classic terms, the shape of evolution is not a tree, but rather a bush; not a ladder, but instead a cone.  (See especially Gould 1987.)  But I would like to draw attention to another important moral of this story: the necessity of obtaining and preserving collections of specimens from all over the world.

In addition to Gould’s original lesson, the rather surprising history of the paleontological study of horse evolution also reveals that you can have what appears to be absolutely everything you need in order to successfully deconstruct the evolutionary history of a lineage, even when looking only at samples from a somewhat limited locale.  And then you can use your subset of fossil specimens in order to construct a morphologically meticulous, environmentally cohesive, theoretically validating, absolutely elegant account of the evolution of the lineage.  But you can still be wrong about it.  The real action might have been happening elsewhere all along.

This is why we need wide-ranging fossil specimens from diverse locales to be excavated, preserved, and made available for study by paleontologists.  When we restrict our sampling to certain areas, we increase our risk of misconstruing evolutionary history in the way that both Huxley and Kovalevsky did.  We also decrease our chance of catching such errors.  Imagine if paleontology had never left Europe, or if the Americans Edward Drinker Cope and Othniel Charles Marsh had not combed the fossil beds of the Western US quite so ruthlessly as they did.  Lately, I have watched delightedly as the story of horse evolution in the Americas is being enriched by, for instance: the use of stable isotopic evidence from fossil horse specimens from throughout South America (Prado, Sánchez, and Alberdi 2011); the genomic sequencing of bone recovered from permafrost in the Yukon (Orlando et al. 2013); and the comprehensive incorporation of Cenozoic specimens from Mexico (Vargas, Bravo-Cuevas, and Jiménez-Hidalgo 2016).

How many other evolutionary stories are waiting to be re-told, made more accurate, filled out, or otherwise enhanced by the incorporation of specimens already stored in museums around the world, or even those not yet collected?  This is something that we lose when specimens are lost: the chance to learn that we were wrong, and the chance to learn how to be right instead.

That’s my first point, about the value of collections.  I’m going to make my second point, about the difficulty of maintaining them, much more succinctly.  I followed the news of the fire at the National Museum in Brazil quite closely, and I was shocked by one particular reaction to the tragedy that I saw exhibited over and over again.  I noticed countless people reacting with variations on the following: “what, they don’t have sprinkler systems in Brazil?”  And it is true that the National Museum lacked a working sprinkler system, and that nearby fire hydrants malfunctioned when they were needed the most.  Museums need functioning fire suppression systems.  But that doesn’t mean disasters like this one are easy to prevent.  I just want to say something quickly right now about how hard it is to properly protect museum collections.

Scientific and natural history museums can hold up to tens of millions of specimens.  Many botanical samples are dried and extremely flammable.  Many zoological specimens are preserved with the help of chemicals like formaldehyde.  Museums often have massive collections of “wet” specimens that are basically dead animals in large, fragile glass jars filled with heavy, flammable fluids (such as alcohol).  Paleontological and geological specimens are often also very heavy—they are basically pieces of rock, some of which are very large.

Now imagine that you were trying to store this stuff yourself—giant piles of it, literally millions of pieces of it.  Normally, you’d want to put the large, heavy, fragile, flammable stuff on the lower floors of your building, right?  But we often ask museums of this type to do a difficult sort of double duty for us: we want them to serve as monuments for public display and education, plus we want them to serve as bastions of scientific research.  Making a museum be and feel genuinely accessible to the public often means putting the open, exhibit-based spaces on the lower floors.  This correspondingly tends to situate the museum’s scientific work and specimen collections on upper floors (where it is hard to protect), or offsite (where it becomes difficult to access and integrate).  One solution can be to build a massive storage facility under the museum itself—but an operation like this can cost millions of dollars, and the space has to be available for development in the first place.

My point here is just that this is not an easy situation to navigate.  It takes serious money, and serious motivation, to properly prepare for and prevent risks to specimen collections—especially from fire.  You have to be willing to spend a whole lot of resources on a bunch of decidedly unglamorous items like underground storage, increased water pressure, load-bearing structural reinforcement, and fire sprinkler and suppression systems.  Finally, note that for many of the items in a museum’s collection, to spray them with large quantities of dry chemicals and / or wet agents is to damage them as specimens.  So, in many cases you’re looking at specialist fire sprinkler and suppression systems—the extra-expensive kind.

Certainly, we must do a better job protecting our specimen collections, all over the world.  We cannot reasonably keep expecting public institutions like our museums to go on successfully serving us, when we are doing so little to care for them in return.  But it’s difficult to see how to make the required changes happen, or where the necessary resources should come from. Museums around the world are facing these difficulties; this is not just a problem in Brazil.

My condolences to the scientific community in Rio.


Darwin, C. 1859. On the Origin of Species by Means of Natural Selection, or the Preservation of Favoured Races in the Struggle for Life. London, UK: John Murray.

Gould, S.J. 1980. Hen’s Teeth and Horse’s Toes. Natural History 89(7): 24–28.

Gould, S.J. 1987. Life’s Little Joke. Natural History 96(4): 16–24.

Gould, S.J. 1996. Mr. Sophia’s Pony. Natural History 105(6): 20–24, 66–69.

Jensen, J.V. 1988. Thomas Henry Huxley’s Lecture Tour of the United States, 1876. Notes and Records of the Royal Society of London 42(2): 181–195. 

MacFadden, B.J. Fossil Horses—Evidence for Evolution. Science 307: 1728–1730.

Marsh, O.C. 1874. Notice of New Equine Mammals from the Tertiary Formation. The American Journal of Science and Arts 7(39): 247–258.

Orlando, L., A. Ginolhac, G. Zhang, D. Froese, A. Albrechtsen, M. Stiller, M. Schubert, et al. 2013. Recalibrating Equus evolution using the genome sequence of an early Middle Pleistocene horse. Nature 499: 74–78.

Prado, J.L, B. Sánchez, and M.T. Alberdi. 2011. Ancient feeding ecology inferred from stable isotopic evidence from fossil horses in South America over the past 3 Ma. BMC Ecology 11(1): 15.

Priego-Vargas, J., V.M. Bravo-Cuevas, and E. Jiménez-Hidalgo. 2016. The record of Cenozoic horses in Mexico: current knowledge and palaeobiological implications. Palaeobiodiversity and Palaeoenvironments 96(2): 305–331.