Malaria and Rome: A History of Malaria in Ancient Italy

by Robert Sallares

  Marchiafava wrote as follows: ‘I may say, at once, on the basis of forty-three years’ experience, that pernicious symptoms arise only in connection with . . . P. falciparum [and not P. vivax].’ He later went on to differentiate P. falciparum from P. vivax malaria by ‘the long duration of the febrile access, viz. from 24 to 36 hours or more, with brief and poorly defined intervals.²⁰ The tertian character of the fever is recognised from the period elapsing between the onset of one access and that of the next, and in pure forms, it can clearly sex per accessionem occupat (interdum etiam vel minus vel plus), neque ex toto in remissione desistit, sed tantum levius est. Id genus plerique medici Ómitrita∏on appellant. Compare the Hippocratic definition of a semitertian fever as a fever which becomes milder on one day, without completely ceas-ing, and then severe again on the following day: Hippocrates, Epidemics 1.2, ed. Littré (1839–61), ii. 606–9: tÏ m†n Òlon oÛ diale≤ponteß: Ø d† trÎpoß Ómitrita∏oß: m≤an koufÎteroi, t∫ ‰tvr7 ƒpiparoxunÎmenoi.

  ¹⁷ Celsus, de medicina 3.8.2: plurimique sub alterutro curantis errore subito moriuntur. Ac nisi magnopere res aliqua prohibet, inter initia sanguis mitti debet, tum dari cibus (Many patients die suddenly as a result of various errors by doctors. And unless there is a strong reason for not doing so, bleeding should take place at the beginning of the illness, and then food should be given to enable the patient to sustain a long period of fever without aggravating it).

  ¹⁸ Cipolla (1992: 57) for Magiotti. Thomas Sydenham regarded blood-letting as very dangerous during acute attacks of malaria in the autumn (Dewhurst (1966: 134–5); Meynell (1991: 127) ). Similarly Palmero and Vega (1988: 346) quoted the Spanish doctor Andrés Piquer as arguing in the eighteenth century that bleeding as a therapeutic method eventually led to the death of patients suffering from malaria.

  ¹⁹ Cato in Pliny, NH 29.7.14.

  ²⁰ Marchiafava (1931: 4, 11), cf. Sambon (1901 c: 348).

  16

  Types of malaria

  be seen that the attacks come on alternate days’. He stressed that all the symptoms of pure infections with P. falciparum are more intense and more severe than the symptoms of pure infections with P. vivax. The obvious similarity to Celsus’ description was noted by Marchiafava himself, and it does not end there. In a passage too long to be reproduced here Celsus then went on to describe quotidian fevers, emphasizing the great range of variation they displayed.²¹ The sheer length of his description of quotidian fevers relative to his description of tertian fevers indicates the great frequency and severity of quotidian fevers in ancient Rome and correlates precisely with the results of the modern research in medicine mentioned above. Celsus’ account was indeed based on an accurate assessment of the relative importance of the various classes of malarial fever. Similarly Marchiafava, on the basis of the experiences of doctors treating cases of malaria from the Roman Campagna in the Santo Spirito (for men) and San Giovanni (for women) hospitals in Rome in the late nineteenth and early twentieth centuries, went on to describe quotidian fevers, irregular intermittent fevers, and subcontinuous fevers. These were all frequently caused by P. falciparum, as was confirmed by microscopic examination of blood smears. He emphasized that the subcontinuous fevers were generally serious, more dangerous than cases exhibiting the classic tertian periodicity. Marchiafava also emphasized that only primary attacks or early recrudescences can become pernicious.²²

  After that, survivors are developing acquired immunity. Consequently people suffering from pernicious malaria often do not yet have enlarged spleens, another classic symptom of malarial infection, noted for example in the Hippocratic treatise Airs, Waters, Places. What are being described in such texts are survivors with a long history of repeated infections, not those killed directly by the disease at an early stage of infection. Splenomegaly is more pro-nounced in cases of P. malariae than in cases of P. falciparum. It is least frequent in P. vivax malaria. A spleen may weigh as much as twenty times its normal weight as a result of malaria. The splenomegaly associated with malaria was already well known to the Romans by ²¹ Celsus 3.3.3–6.

  ²² Marchiafava (1931: 14–20), cf. Marchiafava and Bignami (1894: 231–2) for their views on Celsus. The San Giovanni hospital, near the church of San Giovanni in Laterano, is now Rome’s main accident and emergency hospital.

  Types of malaria

  17

  2. Ospedale di San Giovanni, in Piazza di Porta San Giovanni, is one of Rome’s major hospitals. It formerly received patients with indigenous malaria.

  the second century , and Cato prescribed cabbage as a remedy for splenomegaly.²³

  There is no doubt that P. falciparum malaria was present and well known in all its varied manifestations in Rome at the beginning of the imperial period. It is sometimes suggested that it was a completely new disease in Roman times.²⁴ However, it must be stressed that there is no positive evidence whatsoever for this hypothesis.

  There is not a ghost of a sign in the writings of Celsus, or Galen, or any other ancient Greek or Roman author, that any type of malarial fever was regarded by anyone as a new disease at any time in the period c.500 – c. 500. Both Celsus and Galen had no doubt that they could recognize all the types of intermittent fever ²³ Hippocrates, Airs, Waters, Places 24: ej mvntoi potamo≥ m†n m¶ e÷hsan, t¤ d† data limna∏3 te ka≥ st3sima p≤noien ka≥ ‰l*dea, ån3gkh t¤ toiaıta e÷dea progastrÎtera ka≥

  splhn*dea e”nai (However if there are no rivers and it is necessary to drink stagnant water from marshes, people’s bodies inevitably have swollen bellies and enlarged spleens); Cato, RR 157.7: et si bilis atra est et si lienes turgent; Gilles and Warrell (1993: 54) on the weight of the spleen; H. M. Gilles in Wernsdorfer and McGregor (1988: i. 774) lists other possible causes of splenomegaly.

  ²⁴ Bruce-Chwatt and Zulueta (1980); Zulueta (1973), (1987), and (1994); Karlen (1996).

  18

  Types of malaria

  existing in their own time in works in the Hippocratic corpus dating to the fifth and fourth centuries . The Hippocratic texts are basically very similar to, but less detailed than, the later Roman texts. For example:

  The most acute, most serious, most difficult and deadliest diseases were continuous fevers. The safest and easiest of all, but the longest in duration, was quartan fever . . . acute disease occurs in the fever called semitertian, which is more fatal than the others . . . exact tertian fever reaches its crisis rapidly and is not fatal.²⁵

  Although it would be nice to have more detail, basically this text looks very similar to Celsus, and Galen had no difficulty interpreting it in his commentary on this Hippocratic text.²⁶ It lists continuous fevers as the most dangerous of all (in agreement with Marchiafava), semitertian fevers as the most dangerous of the others (similar to Celsus), tertian fevers as not dangerous, and quartan fevers as the safest of all. Grmek has also adduced other evidence for P. falciparum malaria in the fifth century  with his exquisite retrospective diagnosis of the case of Philiscus in the Hippocratic Epidemics as blackwater fever. In blackwater fever a breakdown of erythrocytes on a massive scale causes excretion of haemoglobin in the urine. This complication of P. falciparum malaria occurred not infrequently in adults in Mediterranean countries in the past.²⁷ The balance of the evidence, pace Zulueta, is that P. falciparum was already present in the fifth century  in Greece. Marchiafava himself had no doubt whatsoever about this and quoted the very same text of Hippocrates quoted above. The origin of the name semitertian was uncertain in antiquity. Galen wrote that he could fill three volumes with what had been written about it previously. This indicates both its importance and its antiquity, since the origin of the term was hopelessly lost. Galen himself interpreted semitertian fever as a combination of tertian and quotidian fevers, although other ancient medical authors may have interpreted it in terms of other combinations of fevers.

  ²⁵ Hippocrates, Epidemics 1.11, ed. Littré, ii. 672–5: ejsi d† øx»tatai m†n ka≥ mvgista
i ka≥

  calep*tatai noısai, ka≥ qanatwdvstatai ƒn t‘ xunece∏ puret‘. åsfalvstatoß d† p3ntwn, ka≥ &∫stoß, ka≥ makrÎtatoß Ø tetarta∏oß . . . ƒn d† t‘ Ómitrita≤8 kaleomvn8 xump≤ptei m†n ka≥ øxva nous&mata g≤gnesqai, ka≤ ƒsti t0n loip0n o˜toß qanatwdvstatoß . . . trita∏oß åkrib¶ß, tacukr≤simoß, ka≥ oÛ qanat*dhß.

  ²⁶ Galen 17A.227–36K.

  ²⁷ Grmek (1983: 409–36); James (1920: 159–62); Stephens (1937: 530–3); Marchiafava (1931: 32–3) on blackwater fever.

  Types of malaria

  19

  It is important to distinguish the problem of the etymology of the word semitertian—a linguistic and philological problem—from the problem of identifying the clinical symptoms described under that heading, a problem in retrospective medical diagnosis.²⁸ Galen rightly formulated the problem in this way:

  the question is not about the reality of the disease, but about the meaning of the name, in relation to those fevers which are called tertian, semitertian, and fevers with a tertian form²⁹

  He then observed correctly that in relation to periodic fevers recurring on the fifth, seventh, or ninth days, which some other ancient medical authors claimed to have observed, the right question to pose concerned not the meaning of the names, but the very existence of these diseases. The uncertainty about the origin of the word semitertian is not a good reason for doubting the equivalence of the symptoms described under that heading with those of a recog-nizable modern disease, pace Jarcho.³⁰ It simply reflects the complexity of the manifestations of the disease, which has already been noted. The testimony of Marchiafava, which has already been examined, shows that Jarcho was quite wrong to suggest that modern doctors cannot recognize the disease that Celsus and Hippocrates called semitertian fever. Celli and the other early modern Italian malariologists and numerous other modern specialists such as Grmek have all made the same identification. Furthermore, the evidence adduced by Jarcho of the failure of early modern authors to recognize semitertian fever is not pertinent, since most of the authors he considered lived in cold climates in northern Europe where P. falciparum malaria could not possibly have been endemic.

  Consequently it would hardly be surprising, for example, if the famous English doctor Thomas Sydenham in the seventeenth century had not recognized a disease which could not have been ²⁸ Galen 17A.120K: tr≤a moi nom≤zw bibl≤a plhrwq&sesqai, cf. 17A.228K for Agathinus’

  discussion of this problem; 7.363K: Ó g3r toi s»nqetoß ƒk trita≤ou ka≥ åmfhmerinoı sunecoıß, n Ómitrita∏on ønom3zomen; 7.369K; Oribasius, Synopsis 6.23: Ø åkrib¶ß Ómitrita∏oß ƒx jsosqenoıß kr3sewß duo∏n pureto∏n ginÎmenoß, trita≤ou diale≤pontoß ka≥ åmfhmerinoı sunecoıß (A genuine semitertian fever is made up of an equal mixture of two fevers, a tertian intermittent fever and a continuous quotidian fever); Aetios of Amida 5.82, ed.

  Olivieri (1950); Marchiafava and Bignami (1894: 88 n. 1). W. D. Smith (1981: 7) commented that the author of Hippocrates’ Epidemics 7.43, 94–6 simply assumed that his audience would know what a semitertian fever was.

  ²⁹ Galen 17A.222K: oÛ per≥ pr3gmatoß, åll¤ per≥ shmainomvnou ka≥ ønÎmatÎß ƒstin Ó

  z&thsiß, oÙß t0n proeirhmvnwn puret0n ønom3zein pros&kei trita≤ouß, Ómitrita≤ouß ka≥

  tritaiofue∏ß.

  ³⁰ Jarcho (1987), contrast Sambon (1901 b, 305–10).

  20

  Types of malaria

  endemic in London in his own time during the climatic conditions of the ‘Little Ice Age’ (see Ch. 4. 5 below), although he undoubtedly did recognize ‘ague’ or ‘marsh fever’, the indigenous English malaria caused by P. vivax (see Ch. 5. 4 below). Even as late as the nineteenth century, Littré in his great edition of the Hippocratic corpus had to insist that the diseases described in it had to be searched for among those present in Greece, not among the diseases of northern Europe.³¹

  Any attempt to trace the history of P. falciparum malaria in Greece before the fifth century  encounters the perennial problem of the shortage of evidence, which bedevils the historiography of archaic Greece. There is no reason whatsoever for supposing that authors such as Hesiod or the lyric poets ever envisaged producing ancient equivalents of Pauly’s Realencyclopädie. Consequently arguments from silence based on such sources are just as worthless in respect of malaria as they are in respect of other problems in Greek history. There is one passage in Homer which has attracted attention. The reference to ‘fever’ (puretÎß, a hapax legomenon in Homer) in such an early source is by itself inadequate to prove that this text is a reference to malaria. Although there is no doubt that specialist medical literature such as the writings of Celsus and Galen does refer specifically to malaria, there will always be room for doubt whether the vocabulary of authors who were not specialists in medicine was used in the same way as doctors used it.³² The meaning of the Homeric text was already debated in antiquity, as shown by the scholiast.³³

  ³¹ Celli (1933: 47); Œuvres complètes d’Hippocrate, ed. Littré, ii. 538–84; Bruce-Chwatt (1976) on the history of the term ague. Pace Jarcho, there did continue to be some recognition of semitertian fevers in literature in northern Europe in the early modern period (e.g. Black (1789: 46, 49).

  ³² The brief reference to an epidemic attributed to ‘bad air’ during the summer in Virgil, Aeneid 3.137–42 raises similar problems of interpretation. There is no detailed description of the symptoms. Livadas (1959: 301) quoted a passage from the ‘Orphic’ Lithika as ‘pre-Hippocratic’ evidence for malaria in Greece in the sixth century , but this poem is now regarded as a late work of the fourth century  (so West (1983: 36) following R. Keydell in Pauly-Wissowa, RE xviii.2 (1942), cols. 1338–41): ej d† puriflegvqwn ‰ter&meroß £ndra qam≤zwn | ∂ kruerÏß m3rptwn puretÏß paradhq»n7sin, | ]† tetarta≤hß p[ma brad», m& pote l&gein | Boulomvnhß, åll’ ajvn, Òp7 pel3s7si, meno»shß, | tÎnde s» g’ j$sqai di’

  åm»monoß åntiac3tou: | o˜toß g¤r puret0n polŸ fvrteroß (If a fever blazing like fire on alternate days regularly attacks a man, | or a fever creeps in while the man is in the grip of an icy chill, | or a quartan fever, never wishing to cease, but always remaining, | brings slow misery to a man as it approaches, | you cure this fever with an excellent agate[?]. | This stone is much better than fevers.) ( Lithika 633–8, ed. Abel (1885) ).

  ³³ Òti ‹pax ƒntaıqa Ø puretÎß, AAext ka≥ Òti puretÏß kur≤wß lvgei, oÛc ¿ß tineß dvcon-Types of malaria 21

  The old man Priam was the first to see him [sc. Achilles] with his eyes, as he hurried across the plain, shining like the star which rises at harvest time [sc. Sirius]. Its rays are conspicuous among the numerous stars in the dead of night, and it is called the dog of Orion by men. It is the brightest star, but its appearance is a sign of evil, for it brings much fever to wretched humans.³⁴

  Nevertheless, if the circumstantial details provided in this passage are approached with an open mind, without any preconcep-tions as to when P. falciparum reached Greece, the whole context is much more suggestive of P. falciparum malaria than anything else.

  The parallel with Achilles, avenging the death of Patroclus, suggests a disease capable of causing very high mortality rates.³⁵ The rising of Sirius (in late July) was definitely associated with malarial environments in Etruria by Tibullus:

  Remain by the water which flows from Etruscan springs, a stream not to be approached during the heat at the time of the dog star, but now second only to the holy waters of Baiae, when the ground thaws during the varie-gated spring.³⁶

  The association with the harvest recalls the name ‘aestivo-autumnal fever’ that was frequently given to P. falciparum malaria by Italian doctors, such as Marchiafava and Bignami in the late nineteenth century. In contrast P. vivax only requires temperatures of 15–16ºC to complete its cycle of development inside the mosquito and has a tendency to cause spring relapses following primary infections in the previous autumn. Consequently
vivax malaria was tai t¶n di3kausin toı åvroß: prÏß g¤r tÏ fqoropoiÏn Ó parabol&. ka≥ Òti deilo∏sin ånt≥ toı deila≤oiß (The word ‘fever’ only occurs once [sc. in Homer]. Moreover it really means ‘fever’, not ‘burning of the air’ as some maintain, for the comparison is with something which causes destruction): Scholia Graeca in Homeri Iliadem [ Scholia Vetera], ed. H. Erbse (1972), v. 268–9. The same pair of alternative explanations is found in Eustathius, Commentarii ad Homeri Iliadem pertinentes, ed. van der Valk (1987), iv. 566, writing in the twelfth century .

  ³⁴ Homer, Iliad 22.25–31: tÏn d’ Ø gvrwn Pr≤amoß pr0toß ÷den øfqalmo∏si, |

  pamfa≤nonq’ ¿ß t’ åstvr’ ƒmpess»menon ped≤oio, | Òß 3 t’ øp*rhß e”sin, år≤zhloi dv oÈ

  aÛga≥ | fa≤nontai pollo∏si met’ åstr3si nuktÏß åmolg‘: | Òn te k»n’ ∞Wr≤wnoß ƒp≤klhsin kalvousi. | LamprÎtatoß m†n Ò g’ ƒst≤, kakÏn dv te s[ma tvtuktai | ka≤ te fvrei pollÏn puretÏn deilo∏si broto∏sin.

  ³⁵ The rising of the dog-star was regarded as an important time in the development of seasonal disease epidemiology by Hippocrates, Airs, Waters, Places 11.