Smallpox, We Hardly Knew Ye*
In an important new study, Duggan et al. sequenced the genome of a 17th-century smallpox virus (Variola major) from a mummy in Lithuania. The results were striking. Perhaps the most important claim of the study is that all known modern strains of smallpox (based on a few dozen globally distributed 20th-century genomes) descend from a very recent ancestor, only some 400-500 years ago. Thus, the Variola virus as we know it, one of the most lethal pathogens in human history, is in some sense a recent nemesis.
The study is a prime example of how genomic evidence is revolutionizing the history of human health and disease. Ancient DNA analysis, for instance, has immeasurably enriched our knowledge of true plague, Yersinia pestis. This enterprise requires the joint efforts of scientists and humanists working across traditional disciplinary boundaries. The new smallpox study makes historical arguments; it begins by citing literature that claims smallpox has been present in “Egypt, India, and China dating back millennia.” That is an accurate representation of what a previous generation held, as expressed in syntheses of McNeill, Fenner, Hopkins, and other literature cited in the study. But, in the kind of scientific article presenting the results of the genomic study, it is not possible to do full justice to the complexity of the historical evidence. Moreover, historians have not been especially helpful in providing accessible synthesis of the latest thinking on smallpox in the historical record, in part because at present there isn’t much of a consensus! Maybe most importantly, though, the new study will provoke us to read the historical evidence differently. The new findings should probably shift the burden of proof or at least suggest a more critical reading of some of the written evidence. It is imperative to get the history right – not only because we ultimately care about the human past, but also for the sake of the science. The historical record provides clues about where and how to look, and it provides confirmation or contradiction that can enrich the physical evidence of the genetic record.
Unfortunately, mining the historical record won’t be easy! Expertise is required in so many languages and traditions that the study of global diseases has to be collaborative. So here I will offer as a resource a few of what I consider the most salient examples from the historical record for smallpox-like viruses from the first millennium. This blog format lets me link to web resources that will allow readers to explore further. I also hope that others may collaboratively add to or detract from what I suggest.
A few quick notes on diagnosis. First, we must have a sense of the symptoms that allow differential diagnosis. Until even recent times, smallpox was confused with measles and chickenpox, especially in its early phases or mild cases.
SMALLPOX typically presented with a prodromal fever lasting 2-4 days, often attended by severe back pain, before the onset of rash; its characteristic rash included lesions over the whole body but especially the extremities, including the face and often the palms and soles. The rash arose in all parts of the body simultaneously (not in crops) as large, deep pustules that felt hard and deeply embedded in the skin. The rash resolved slowly over 2-3 weeks, before scabbing off and leaving scars. Mortality was often high, with 10-30% case fatality rates. In epidemic contexts, a minority of cases presented hemorrhagic symptoms, in either an early or late stage form; these were nearly always fatal. In short: deep pustules, developing simultaneously, covering the whole body though especially the extremities rather than the trunk/back, a lengthy course, potentially lethal.
MEASLES typically presented with a prodromal fever lasting 2-3 days, accompanied by a cough, coryza, and conjunctivitis. A red, maculopapular rash spreads from the head to the entire body, lasting around 8 days. Unlike smallpox, the measles rash is not characterized by hard pustules extruding from the skin. Measles mortality can depend on nutritional status and comorbidity but without modern treatments could be high, though not as high as smallpox, ca. 10%
CHICKENPOX typically presented with a fever and rash simultaneously. Lesions spread over the body but not the palms and soles. The vesicles are smaller and shallower than smallpox and resolve more quickly, within 7-10 days. Case fatality rates are low, although they could have been higher in premodern populations with poor nutritional status and high comorbidity. But there is no mistaking that chickenpox is far the mildest of these three.
Second, we must approach retrospective diagnosis with an abundance of caution. It is surpassingly difficult to identify the causative agent of an ancient mortality event from records of the pathology of a disease. Of course, it’s not always impossible: our two main eyewitnesses to the Justinianic Plague, Procopius and John of Ephesus, offer vivid descriptions that effectively pin-pointed Yersinia pestis as the pathogenic agent (though see closing thoughts for some ways in which, even still, the genomic evidence is meaningfully additive and not merely confirmatory). But sufficiently detailed and accurate descriptions providing just the right kind of observations are rare. Many impediments could stand in the way. The lack of germ theory sometimes made it hard to distinguish among specific agents in a world where there was always a welter of diseases; literary tropes were prominent; our observers are often not medical experts.
So, beyond the symptoms that may be described by contemporaries, historians of disease have increasingly paid attention to clues about the epidemiology – the population-level behavior – of specific pathogens. Further, we increasingly have some sense of the phylogeny and phylogeography of specific pathogens – family trees that help us establish the evolutionary map, in time and space, of particular microbes. And best of all is direct sequencing of archaeological genomes, such as we now have from the 17th-century sample of smallpox. Thus, the study of historical pathogens depends on analysis of (1) pathology; (2) epidemiology; (3) phylogeny; and (4) archaeological genomics. Obviously, what we most need now are even older putative smallpox specimens from the relevant historical contexts.
What follows is an overview of some of the main evidence for smallpox-like viruses from before AD 1000. I won’t burden the reader by saying smallpox-like or using scare quotes – suffice it to say, whether any of these diseases were smallpox or a relative of smallpox is the question. The European evidence for the late middle ages/early modern period is brilliantly overviewed in a contribution of Carmichael and Silverstein.
The mummy of Ramesses V (ruled 1149-45 BC) was discovered in 1898. Donald Hopkins believed that lesions on the face suggested smallpox and popularized the view in his widely read book on the history of smallpox. This is not much evidence, nor is it very convincing.
For the Indian evidence, I am indebted to an insightful and balanced study of Ralph Nicholas. There are four testimonies worth noting – two from some of the earliest Sanskrit medical compilations and two from early medieval medical texts.
The Suśruta Saṃhitā and Caraka Saṃhitā are two of the most important and earliest compilations of Indian medical literature. Both texts have complex transmission histories and are difficult to date definitively, as they preserve layers of ancient tradition as well as later accretions. The Suśruta Saṃhitā may have strata going back to the mid-first millennium BC, but it may have taken its present shape in the first centuries of this era. Similarly, the Caraka Saṃhitā exists in what is likely a fourth-century redaction.
In the Suśruta Saṃhitā (Chapter 13) a number of diseases that include skin pathologies are catalogued, including masūrikā, which means smallpox in the later tradition. The reference here is brief and uncertain: painful pustules covering the entire body as well as intense fevers, but no indication that the condition was fatal. In the Caraka Saṃhitā (6.12), there is an even more extensive discussion of a variety of pustular diseases, but only a passing reference to masūrikā, which is characterized as a disease with pimples the size of seeds spreading over the entire body. It is doubtful that the text is referring to smallpox, and it has been inferred from the absence of evidence here that smallpox was not familiar in India in the fourth-century AD (see the acute observations of Valiathan, The Legacy of Caraka, xxix-xxxi).
The evidence of two later medical texts is more significant. The Aṣṭāṅgahṛdayasaṃhitā, written by Vagbhata in the sixth or seventh century, describes masūrikā as a deadly disease characterized by a rash with “coral globules.” Most important of all, the Madhava nidanam, written by Madhava-kara in the early eighth century, devotes a chapter to masūrikā, displaying knowledge of smallpox as well as other diseases apparently including measles and chickenpox. The observations are truly fascinating. I am entirely dependent on the 1987 English translation by K.R.L. Gupta. In a chapter on masūrikā, Madhava-kara describes what seems to me like a cluster of diseases with skin pathologies – smallpox likely among them. Here, the disease presents first with a fever, including pain particularly in the back. Restlessness, itching, redness in the eyes and face, and cough occur. The pustules were dark-green or red, dull, hard, and rough. They “ripened” slowly. The pustules could take different colors and be different sizes. Sometimes the disease was accompanied by diarrhea and ulcers in the mouth. In other cases, the pustules were black and flat and full of discharge which was plentiful and fetid. It could be fatal. Madhava-kara was also familiar with “measles,” or a disease with “the eruption of small red papellae, like the roots of hairs; these appear all over the body” and were accompanied by cough.
As with Rhazes, discussed below, it seems in the case of Madhava-kara that he observed a variety of diseases including smallpox, as well as possibly measles and chickenpox. Even if he struggled to differentiate them, the observation of an acute deadly disease with fever and back-ache followed by slow developing extrusive pustules points to the presence of smallpox in India by the early eighth century.
In China, the main evidence which is cited is a passage of the well-known fourth-century writer Ge Hong (sometimes transliterated as Ko Hung). His work Chou hou pei chi fang, “Handy Therapies for Emergencies,” has not been translated into any European language, unfortunately. The work was written around AD 340 and redacted around AD 490 by Thao Hung-ching. I quote an excerpt translated by Joseph Needham (italics his).
“Recently some people have suffered from seasonal epidemic eruptions which attack the head, face and trunk. In a short time they spread all over the body. They look like fiery [red ] boils, all containing a white fluid. The pustules arise all together, and later dry up about the same time. If they are not treated immediately, many of the more severely afflicted patients will die in a few days. Those who recover are left with purplish or blackish scars, the colour of which takes years to fade. This is due to a severely poisonous chhi. People say that it first appeared from the West in the fourth year of the Yung-hui reign-period, and passed eastwards, spreading all over the country… In the mid Chien-wu era [our soldiers] caught it when attacking marauders at Nan-yang; for this reason one of its names is still ‘marauders’ pox’.”
The symptoms are certainly consistent with smallpox, but not really detailed enough to permit a confident diagnosis. More work on the Chinese evidence is a desideratum.
Around AD 165, the Roman Empire was rocked by a pandemic mortality event known as the Antonine Plague (so called after the family name of the emperors Marcus Aurelius and Lucius Verus). It marks a point of rupture in Roman history, and ancient historians have increasingly identified the pandemic itself as a kind of turning point. Much more could be said, but I’ll briefly sketch some of the main outlines, because what is of interest here is that Roman historians working today have reached a virtual consensus on the point that the agent of the mortality was smallpox. Until there is positive genomic evidence, it is important to keep an open mind. There is a real chance that the agent of the pandemic was not smallpox, and until genomic evidence confirms or excludes that hypothesis, all possibilities need to remain in consideration. However, I do think the reasons for believing the biological agent of the Antonine Plague was smallpox or a near ancestor of it are ultimately convincing.
The disease event was clearly exceptional in its magnitude and significance. While the Roman Empire was an insalubrious place, and its towns and villages subject to the ravages of epidemic mortality, there is absolutely no evidence for pandemic disease in the Roman world (and arguably the ancient world altogether, although the Plague of Athens in the fifth-century BC is the best candidate) prior to the Antonine Plague. In other words, there were local and possibly regional disease events, but not continental-scale disease outbreaks before this one. It takes an exceptional pathogen to do what this one did. Major mortality is attested in Arabia in the AD 150s — quite remarkably, both in Roman historical sources and in Arabian inscriptions — that probably reflect the advance of the disease toward Roman borders. By the mid-160s, the disease was inside the empire. It struck Egypt and the Levant, Asia Minor and Greece, Italy and central Europe, North Africa and Britain — all in the space of a few years. It left a deep mark on the historical and annalistic traditions, which testify to an unprecedented event. More remarkably, scattered evidence from inscriptions and papyri confirm an unusually widespread and devastating mortality. Roman history is alive and well, and we keep finding evidence for the Antonine Plague as well as noticing evidence we had missed. There was simply nothing like this before. Historians debate the death toll and social impact of the disease. In my own work, I argue for a mortality of no more than 10% of the imperial population. Some estimates are even higher, but a disease event taking off some 7 million people would probably have been the largest single mortality event in absolute terms up to that point in human history. Moreover, the consequences reverberated directly and indirectly for decades. Many scholars, including myself, argue that it caused a major shift in factor prices, and it evidently strained the empire’s ability to recruit soldiers.
Now, one thing that makes the Antonine Plague even more remarkable is that the greatest doctor of the ancient world, Galen (sorry, Hippocrates), was a contemporary. His own biography was shaped by the course of the pandemic. He was in Rome when it erupted, hurried for obscure reasons back to his native Pergamum as it approached, and was quickly summoned by the emperor Marcus Aurelius. We so badly wish he had written a tract on the pandemic, but instead we have to chase down the bits of things he said about it in his voluminous writings. He mentions the Antonine Plague a couple of dozen times, but mostly just to refer to it as “the great plague.” To Galen the etiology of the pestilence was an excess of black bile, literally a “melancholy.” Judging by his scattered notes on the disease, in Galen’s mind its attendant symptoms were fever, a black pustular rash that raised above the skin and eventually scabbed and scarred, ulceration deep in the windpipe, and black or bloody stools. In Galen’s paradigm, those who had a “dry” constitution had the greatest chance of surviving an infection. Galen also believed that the “crisis” came around the 9th day.
Galen believed that the fever putrefied the blood of the victims, and that nature forced an ash-like remnant through the skin. “There was no need of drying medications for such exanthemata [extrusive pustules] for they spontaneously existed in the following manner: in some, in whom there was also ulceration, the surface fell off, which they call scabs, and henceforth what remained was already close to health, and after one or two days scarred over. In others, in whom there was not an ulceration, the exanthem was rough and itchy, and fell off like something scaly, and from this all patients became healthy.” In his treatise On Black Bile, Galen described the black pustules covering the entire body that then dried and fell off like scales, sometimes many days after what Galen considered the “crisis” or turning point of the disease. These clinical observations describe the course of vesicular and then pustular lesions that characteristically scabbed off, leaving scarred but no longer pathological dermis in their place.
For those who wish to consult the primary sources themselves, the two most important passages are De methodo medendi 5.12 and Atra Bile 4. There is a highly valuable treatment by Littman and Littman 1971.
The pathology described by Galen is strongly suggestive of smallpox. We certainly wish he were more specific about the rash (did it concentrate in the extremities? arise simultaneously across the body? feel like inset beads within the skin?). Nonetheless, a highly fatal disease, characterized by an extrusive rash with a lengthy course, points to the virus. The crisis on the 9th day is highly consistent with the course of modern smallpox, if measured from the appearance of the first symptoms. The malaise, diarrhea and occasionally hemorrhagic presentation are all consistent if not diagnostic. We should be open to all possibilities until there is genomic evidence, including more virulent forms of diseases like Varicella zoster virus (chickenpox), but those require a leap of the imagination (could an ancestral herpesvirus known to us in such mild form have caused a deadly intercontinental disease event? It seems doubtful). Measles is the best alternative candidate. It’s deadly and famously communicable. However, Galen does not emphasize respiratory symptoms, and the measles rash does not extrude, scab, and scar the way that smallpox does. Moreover, molecular clock evidence for measles as well as some insightful new work (see below) is providing a more compelling origins story for measles, that starts a few hundred years later.
I say more in forthcoming work, but the pathology, epidemiology, and what we know of the phylogeny of orthopoxviruses at present are all consistent with the hypothesis that the Antonine Plague was smallpox, in some form. At the very least, in the present state of the evidence, I think that is the most likely candidate.
Eastern Mediterranean, AD 312-3
After the Antonine Plague, there is little evidence for endemic or epidemic smallpox in the Roman world. One possible exception is an episode of famine and pestilence noted in the Church History of Eusebius (at 9.8). The event is imprecisely located in the cities of the eastern emperor. The disease was characterized by “a malignant pustule, which because of its fiery appearance was known as a carbuncle. This spread over the entire body, causing great danger to the sufferers; but the eyes were the chief target for the attack, and hundreds of men, women, and children lost their sight” (tr. Williamson). In his comprehensive catalogue of famine an pestilence in late antiquity, Dionysius Stathakopoulos identifies smallpox as the agent of this mortality. The theory is consistent with the description, but the evidence is obviously too limited to say with any kind of confidence. The word translated as “pustule” just means “wound” or “sore,” and the description is far more uncertain than Galen’s.
The only other epidemic that Stathakopoulos ascribes to smallpox broke out in Edessa in AD 494-5 and 496-7; the timing is obscure (One continuous event? Two related events?). We are reliant on a single source, the vivid text known as the Chronicle of Ps.-Joshua the Stylite, at sections 26 and 28 (online here). The affliction involved swellings and tumors and pustules over the entire body, including the soles of the feet and palms of the hands, with scarring in the aftermath. No great death toll is remarked in this first wave, which was followed by a deadlier pestilence amidst a famine, which may or may not have involved the same pathogenic agent. While we should always be wary of single-source accounts, and while Pseudo-Joshua’s views are colored by his religious perspective (specifically, God made manifest on the outside of the body the corruption of the people on the inside), the author of the Chronicle was an eye-witness who reports vivid details, and the presence of sores on the palms and soles is an important differential symptom of smallpox, as is scarring. This is an important possible testimony for epidemic smallpox. [I thank Ana Duggan for thoughtful suggestions].
Gregory of Tours was a Gallo-Roman bishop and historian; he is by far the most important source for western Europe in the sixth century. For the study of disease, his literary corpus is a really remarkable source. Not only does it vividly describe several outbreaks of bubonic plague in the west, he also observes other events that might be ascribed to smallpox and/or measles, as well as several major epizootics. Not all of these events are found in his History, as Gregory was the author of several miracle collections and hagiographies. I discuss his observations of plague in forthcoming work, but let me here try to clarify the main other passages of interest.
In the History, Gregory describes an epidemic that struck “the whole of Gaul” (translations from the History are Thorpe’s) in AD 580 at Hist. 5.34 (Latin here and an older English translation here). He called it a “dysentery.” Its symptoms included high fever, vomiting bile, pain in the back, neck, and head. “The country-folk imagined that they had boils inside their bodies; and actually this is not as silly as it sounds, for as soon as cupping-glasses were applied to their shoulders or legs, great tumours formed, and when these burst and discharged their pus they were cured.” The epidemic started in August and carried off children first. In 5.36, the body of one victim “became so black that you would have thought that it had been placed on glowing coals and roasted.”
Gregory also describes another disease outbreak that ravaged Gaul in AD 582. This event is mentioned in the History (6.14), where he briefly notes “a whole series of malignant diseases, the main symptoms of which were boils and tumours” (i.e., pustules and vesicles; I note briefly that the critical edition actually reads valitudinis variae milinae [rather than malignae, as in some manuscripts] cum pusulis et vissicis and suggests that milinae comes from milium, little seeds of millet). In his Miracles of St. Martin, 3.34, Gregory describes the same outbreak in more detail as it was experienced in his town, Tours. I quote Van Dam’s translation (Saints and Their Miracles in Late Antique Gaul): “It included extreme weakness, so that a man who was seized by a high fever broke out completely with blisters and small pimples. The blisters were white and firm without any softness, although they did produce a sharp pain. Once these blisters became ripe, if they popped and began to discharge, then [people’s] clothes stuck to their bodies and the pain increased more severely… When the wife of Count Eborinus was afflicted by this plague, she was so [completely] covered with these blisters that neither her hands nor any other part of her body remained without blemish; even her eyes were obstructed and covered with these blisters… Soon her fever vanished, the blisters painlessly disappeared, and she was cured.”
Finally, we should note that Gregory (e.g. at Hist. 6.31, 6.44, 10.30, and the Miracles of Martin 6.32) and some of his contemporaries mention an unusual number of livestock plagues. In a recent study considering both historical and molecular evidence, Timothy Newfield makes a very intriguing case that some of these sixth-century pestilences were caused by an ancestral morbillivirus, closely related to measles and Rinderpest virus. I find it plausible, especially in the case of the events of AD 582. While Gregory’s descriptions of bubonic plague are diagnostic, the events of AD 580 and 582 are simply more uncertain. The former could be smallpox, the latter could be measles, but greater certainty is impossible. (See also p. 23 of Newfield’s article for some possible 10th-century references to smallpox that I don’t discuss here).
Plague of AD 665
Arabic sources record an outbreak of smallpox during a military campaign inside Asia Minor in AD 665. Conrad (Plague in the Early Medieval Near East, p. 249) lists the complete sources, e.g. al-Mas’udi. However, beyond using the Arabic word for smallpox, al-Judari, there is no detailed information which might add depth or confidence to the diagnosis.
Aaron (aka Ahrun) of Alexandria
An important sequence of testimony derives from the work of the 9th-10th century Persian medical scholar Rhazes, discussed immediately below. Here we pull out for special notice the fragments of Aaron, a doctor who lived in Alexandria in the first half of the seventh century. Alexandria in the seventh century was a pretty vibrant place (see the Life of John the Almsgiver), in some respects the last outpost of classical civilization. The work of Aaron’s contemporary, the doctor John of Alexandria, gives some sense of the vitality of medical scholarship in this setting. Aaron wrote a medical collection (a pandektēs or syntagma) in 30 books. It is lost except for quotations, but it was translated from Greek into Syriac and from Syriac into Arabic (see here and Ullmann, Die Medizin im Islam, 87-9). His work was well known to Rhazes. It is worth singling out Aaron because he may be the first person on record to describe both smallpox and measles. In the passages cited in the al-Hawi of Rhazes, Aaron describes symptoms of what are named in Arabic “al-Judari” and “al-Hasbah,” which became the terms for smallpox and measles respectively. We would love to know the underlying Greek of Aaron’s original. In the passages preserved in Rhazes (here), Aaron describes “smallpox and measles,” but it is not entirely clear that he distinguishes these as distinct diseases. He does not offer a differential diagnosis, but does seem to focus on al-Judari (smallpox). For both smallpox and measles, he characterizes the symptoms as an acute fever with headache and redness in the eyes, with eruptions that begin around the third day. The smallpox mostly attacked children. Pustules that were greenish or violet foretold the worst.
Abū Bakr Muhammad ibn Zakariyyā al-Rāzī, aka Rhazes, was a Persian physician of the late 9th and early 10th centuries who moved between Baghdad and his native Rey, near Tehran. The first thing to note about Rhazes is that he wrote extensively about the acute pustular diseases al-Judari and al-Hasbah. He is commonly credited as the first physician to offer a differential diagnosis of smallpox and measles, including a treatise dedicated to the subject, Al-Judari wa al-Hasbah. His commonplace book (Al-Hawi) preserves the notes and observations of a number of other medical writers, from Aaron of Alexandria down to his own time. Thus, we possess a string of testimonies about smallpox and measles from the eastern Mediterranean and Near East from the 7th-10th centuries. In 1848, William Alexander Greenhill (just your ordinary sanitary reformer / Arabist) helpfully translated the bulk of Rhazes’ work on smallpox and measles.
We could ask a lot of rich questions about what it could have meant for Rhazes to think that there were two different diseases with distinct etiologies and pathologies. It is fair to question whether everything he observed and described really pertains to these two diseases and these only. He could have observed Varicella zoster virus (chickenpox), Rubella, or some kinds of Erysipelas infections. What I would focus on here, though, are some of the salient features of the clinical observations, especially those that indeed point in the direction of smallpox. I offer the page number in Greenhill.
- Hardly anyone escapes it, especially children, especially male children. It seems to be an endemic disease of childhood (p. 28-9).
- Symptoms include acute continuous fevers (p. 34) and severe back pain (p. 34). Rhazes claims that back pain is more characteristic of smallpox than measles.
- Additional symptoms (p. 44) include redness in the eyes, violent headache, a strong pulse, heavy breathing, and “red and turbid urine.”
- In bad cases, pustules break out in the eyes (p. 51).
- The pustules can be hard and warty (p. 56) and/or filled with fluid (p. 57).
- When the pustules dry, “scabs and dry eschars still remain” (p. 59).
- The diseases (each of them) exist in more mild and more fatal form (p. 71).
- Measles is worse than smallpox, except in the eyes (p. 92).
- And finally, an observation so important that I quote it in full (p. 113): “The difference between the two I have found to be, that the Measles are red, and appear only on the surface of the skin, without rising above it, while the Small-pox consists of round eminences. When these eminences appear, fix your attention on them, and if you are in doubt to the disease, do not express any opinion about it for a day or two; but when there are no eminences, you must not give your opinion that the disease is the Small-pox.” And elsewhere in al-Hawi (p. 121) Rhazes quotes another physician (Abdollah ibn Bukhtishu?) to the same effect, that in the case of smallpox the pustules are raised, like berries, whereas in measles they are level with the skin.
I believe the foregoing evidence strongly supports the conclusion that doctors from seventh-century Alexandria to the 10th-century Near East were familiar with a relatively mild, endemic form of smallpox.
In sum, the historical evidence for smallpox in the first millennium ranges from highly speculative to quite plausible. The most likely testimonies include the evidence for the Antonine Plague, the tradition connecting Aaron of Alexandria and Rhazes, and the 6th-8th century Indian medical texts. We should note a crucial difference in these testimonies: the Antonine Plague was a massively deadly pandemic, striking across three continents in the space of a few years and carrying off millions of victims. The disease known to Aaron and Rhazes and Madhava-kara was an endemic disease. That is a crucial difference that will ultimately need to be accounted for.
How could the historical evidence fit with the picture emerging from the new genetic evidence? Much remains uncertain at this point. But whatever the answer, it is likely to draw from a new paradigm that is starting to emerge in the history of human disease, one that appreciates the recent evolutionary volatility of major human pathogens. Take the study of the plague bacterium, Yersinia pestis, the causative agent of three enormous pandemics. Y. pestis has acquired, and lost, crucial virulence factors in the last few thousand years that have shaped the course of its history. We have recently learned that it afflicted Bronze Age populations across central Eurasia and had most of the genetic package that made it such a deadly disease in later history. The lineage that caused the first pandemic, known as the Justinianic Plague, had acquired a deadly mutation in the code for one of the bacterium’s most important virulence factors (known as pla). Then, we now know, this branch went extinct, while a closely related lineage persisted in central Asia, only to emerge again and cause the Black Death. In short, the evolutionary history of a major pathogen has been a rocky ride, with twists and turns. Contrary to a stubborn misconception, diseases do not uniformly evolve toward less virulent forms. Like any organism, pathogenic microbes evolve in whatever way contributes to their reproductive success, in the conditions they find themselves. The explosion of human populations in the last few millennia, and their increasing global connectivity, has created a remarkable environment for pathogens, and their evolutionary development is correspondingly diverse and volatile.
All of this raises the question for historians of what we mean when we speak of “the” smallpox virus. Defining a virus species is never simple. The last few thousand years – the “historical period” – is a long time in the evolutionary history of a virus like smallpox, an obligate human parasite with no known animal reservoirs. It is probable that the organism has had an eventful history over the last few thousand years and that we only see the broadest outlines of this history. Before the paper of Duggan et al., the most interesting paper on the history of smallpox looked at the phylogeny of the orthopoxviruses. The closest known relative of human smallpox (at present) is Taterapox virus, an orthopoxvirus that infects the naked-sole gerbil, a rodent of the African savannah. Babkin and Babkina argued that smallpox evolved in Africa, 2-4 thousand years ago, from an ancestral rodent virus. (It is worth noting that they derive a different estimation of the mutation rate by focusing on “central” and more conserved regions of the smallpox genome.)
The history of smallpox, then, might look something like this. The disease and anything like it was unknown in ancient medical literature, such as the Hippocratic corpus. It evolved from an ancestral rodent orthopoxvirus – host currently unknown – a few thousand years ago in Africa to become an obligate human parasite. Given the extensive connections across the Red Sea, the Indian Ocean, and the Persian Gulf, it spread into Eurasia. It was extremely communicable and extremely lethal, causing the Antonine Plague. *If* the Antonine Plague was the global debut of smallpox: (1) It could have evolved into the milder, medieval form known from Aaron, Rhazes, et al. (2) It could have been a virulent lineage that went extinct. Or, (3) It could have maintained an obscure profile before re-emerging as the lethal, modern, epidemic form. Other reconstructions are conceivable, too, but whatever the case, refining this story will require the combined efforts of historians and geneticists and the careful analysis of both kinds of evidence. Smallpox is extinct, but our understanding of it continues to evolve.
*I can’t properly acknowledge the debt of my colleagues and friends who have helped guide me, in several cases through languages and traditions that are unfamiliar territory for me. Deonnie Moodie, Garret Olberding, Jack Tannous, Scott Johnson, and Ann Carmichael deserve special thanks.
 Galen’s observations are at Praes. Puls. 3.4 (9.357K); Atra Bile 4 (5.115K); Hipp. Epid. 3.57 (17a.709K); Simp. Med. 9.1.4 (12.191K); Meth. Med. 5.12 (10.367K); K17a.741; 17a.885K; 17a.709K; 17a.710K; 17b.683K; 12.191K; 19.15, 17-8K.
 Galen, Meth. Med. 5.12 (10.367K), tr. Johnston and Horsley.
ἕλκος δὲ ἦν φερωνύμως τοῦ πυρώδουςἕνεκεν ἄνθραξ προσαγορευόμενον—ἐπιφορά, ὃ καὶ καθ’ ὅλων μὲν ἕρπον τῶν σωμάτων σφαλεροὺς ἐνεποίει τοῖς πεπονθόσι κινδύνους, οὐ μὴν ἀλλὰ καὶ κατὰ τῶν ὀφθαλμῶν διαφερόντως ἐπὶ πλεῖστον γινόμενον μυρίους ὅσους ἄνδρας ἅμα γυναιξὶν καὶ παισὶν πηροὺς ἀπειργάζετο.