Few days following the death of rats,

Men pass away like falling walls.

Shih Tao-nan (1792)

The Unholy Trinity: Bacterium, Flea and Rat

In part 1, I discussed how, historically, early modern Europe coped with the plague. The Ring of Fire (RoF) brought physicians, scientists, engineers, books and tools from the year 2000—albeit only what might be found in Mannington, West Virginia—to 1631 Germany. These provide new insights as to how to prevent or even cure the plague, although implementing those methods on a continental scale with the existing infrastructure will be a formidable and time-consuming task.

To achieve effective protection, we have to understand what causes the plague, and how it is transmitted to humans.

The Pathogen

Plague is caused by a bacterium, Yersinia pestis. The organism has a dual life cycle, with differences in body temperature causing different genes to activate in insect hosts than in mammalian hosts. (Abbott 5). In hibernating rodents, the reduced body temperatures may inhibit the bacterium’s virulence as well as the host’s immune response, so the bacterium can “overwinter” in such a host. (46).

To infect humans, just one to ten organisms are needed if entry is by oral, intradermal, subcutaneous or intravenous routes and 100–500 for infection by inhalation. (CIDRAP). It doesn’t form spores, and it’s sensitive to sunlight and heat (Rollins; Pollitzer2, 99ff), so it survives outside a host for long periods only under special conditions. Viable Yersinia pestis have been obtained sometimes from soil (e.g., contaminated rodent burrows).

Biting Insect Hosts

The principal insect hosts are fleas. Their eggs are laid on the host but fall off, usually dropping into the animal or human’s bedding material. They hatch in 2–12 days. The larval stage lasts 1–20 weeks. The larvae eat old skin, hair and other organic matter; they especially enjoy “flea dirt” (adult flea feces) because of its blood content. The larvae spin cocoons and exist as pupae inside for a week to nearly a year.

The adult flea is about 5% of the total flea population. (Fitzgerald 516). The lifespan of the adult flea is dependent on the species and how often it is fed; adult human fleas fed blood daily can live over 500 days, and rat fleas about 100 days. (Pierce 363). A starved Indian rat flea reportedly can survive up to 29 days. (Cipolla/PHMPR 63). Adult fleas can survive on grain debris for three months. (Pollitzer7, 298).

The flea acquires the bacterium as a result of biting an infected host, and transmits it by biting a second host. The human flea (Pulex irritans ) is a poor vector. (Pollitzer7, 294ff). However, there are several species of animal flea in Europe and Asia that if hungry enough will bite humans. (240). For rodent-human transmission, the Indian rat flea (Xenopsylla cheopis ) was the principal villain in the Old World, with the European rat flea (Ceratophyllus fasciatus ) as runner-up. For rat-pet-human transmission, there are the dog flea (Ctenocephalus canis ) and the cat flea (C. felis ). (Pierce 363ff).

Depending on the species, the adult flea may live on the host (as for X. cheopis ), or it may live in nesting material nearby and “dine out” (as for P. irritans ). (Hall 24).

Inside the flea, the plague bacilli multiply, and in some cases form colonies that partially or completely block entry of later blood meals into the midgut. That means that the blood can’t be digested, and consequently the flea begins to starve. The good news is it dies after 3–5 days without food. (Abbott 5; Pollitzer7, 288). The bad news is that it becomes a very aggressive biter. It also tries to clear its gastrointestinal tract by regurgitation, which transmits the bacteria (up to 11,000 to 24,000 in one bite) to the next mammalian victim. (Eldridge 393, Fagerlund 98, Abbott 8).

Foregut blockage occurs readily in the Indian rat flea, and rarely in the human flea. (Pollitzer7, 269). In the European rat flea, it occurs perhaps 12–16 days after the flea is infected. However, according to some authorities, the latter is capable of “early phase” (pre-blockage) transmission, as early as one day after infection, at equal efficiency, even if it isn’t biting as frequently. (Abbott 5). Others insist that “transmission from unblocked fleas is exceedingly rare.” (Perry 51). It’s possible that they are important only when making a “mass attack.” (Pollitzer7, 270).

Fleas reproduce rapidly. Heloise (178) says that “under good conditions 10 fleas can produce a quarter of a million new fleas within 30 days.” The warmer the weather, the less time it takes for a flea to go from egg to adult; 17 days at 89.6oF versus 40 days at 69.8oF. (Vail).

Ticks and lice have been shown to be capable of transmitting plague between various mammals under experimental conditions. (Abbott 27). There is a suspicion that they may play a role in large epidemics.

Mammalian Hosts

Natural Plague Reservoirs

Scientists believe that plague is maintained, between outbreaks, in large colonies of rodents that are infected by but relatively resistant to the disease, such as deer mice and voles in California, or marmots and great gerbils in Asia.. Under circumstances that are poorly understood, but probably include high rodent and flea population densities and territorial overlap with more susceptible rodents, the plague (or more precisely the flea) jumps to a susceptible rodent host such as prairie dogs, ground squirrels, chipmunks and wood rats. (Abbott 38, 44). Even in susceptible hosts, the disease may remain endemic if the flea density is low enough; for prairie dogs, the threshold is supposed to be between 12 and 19 fleas per animal. (43). Unfortunately, we know next-to-nothing about plague “ecology” in seventeenth-century Europe.

In the seventeenth century, the three most important wild animal reservoirs of plague were the ones in the Himalayan foothills of Burma, in the lake region of central Africa, and on the Eurasian steppe. (McNeil 111). There were probably natural foci of plague in Europe, too, but I have not seen these identified. In the modern world, the closest foci are around the Caspian Sea. (WHO).

Rats, Humans, and Pets

While many rodent species are infected by Yersinia pestis, from a human health standpoint the most important rodents are rats. In the seventeenth century, the chief rat vector was the black rat (Rattus rattus). It is aptly called the roof rat (it climbs well) and the ship rat. It has 3–9 litters per year, and 4–8 young per litter. (Hendrickson 84).

In the eighteenth and nineteenth centuries, it was gradually and partially displaced by the brown rat (Rattus norvegicus), a larger and more aggressive rodent with a propensity for burrowing. The latter is dominant in temperate climates and in urban areas. It has 3–7 litters of 8–10 apiece. For both species, an average of 20 young a year are weaned. (Id.)

For every rat you see scurrying about at night, there are probably dozens you don’t. In 1901, on a county estate of 2000 acres, about 32,000 rats were killed within five years by the owner’s crew, and another 5,000 by the tenants. On a 400 acre farm, when the barn was emptied, over 1400 rats were killed. (Forbush 13).

Both black and brown rats are capable of transmitting plague to humans. (Orent 257). The more infected rats, the greater the risk; in one Indian village, human cases occurred in 3–4% of the houses in which one plague rat was found, and 28% of those where more were found. (Pollitzer9, 133).

With individual exceptions, rats are not resistant; plague is just as much of a calamity for them as it is for us. And so infected fleas will cycle rapidly through the rat population, giving them more reason to jump to us. Urban epidemics were often preceded by numerous rat deaths, sometimes taking the form of “rat falls” (especially in buildings with thatched roofs). (Abbott 32). A plague-infested rat is more likely to come out into the open, and is also more readily killed. (Simpson 363).

Because rats live in association with urban populations (providing opportunities for rat-to-human transmission) they are the most important animal host from an epidemiological standpoint. However, in rural areas, other rodents, such as squirrels, may be significant. Cats, and to a lesser extent, dogs can acquire plague by ingesting rodents. (Perry 53). Cat and dog fleas are poor vectors, but cats and occasionally dogs can develop pneumonic plague and transmit it to humans. (cats, Bagle; dogs, Wang).

Because humans are not the only hosts, widespread vaccination does not create a herd immunity effect. The organism will persist in the animal populations and these can re-infect humans.

There is often a seasonal pattern to plague. In northern Europe, cold weather drives the rats underground, away from humans, and so even if the rats die, their fleas are less likely to find a human host. The fleas also hibernate. Thus, human plague cases become less frequent come winter. Come spring, the flea eggs hatch, the fleas become infected, and the rats venture above ground. (Chase 110, 161).


There are two major modes of transmission to humans, by bite from an infected flea, and by inhalation of respiratory droplets from an infected humans.

By bite . For transmission by bite, there are two steps, mammal to flea, and flea to human. The mammal is usually an urban rat, and it need not be alive to be dangerous; a corpse may bear living, hungry fleas.

The volume of blood ingested by the flea is very small (0.03–0.5 microliters) (Perry 51), and the minimum infectious dose for a flea surprisingly large. Hence, to infect the flea, the bacterial level in the mammal’s blood (bacteremia) must be extremely high: 100,000,000 organisms per milliliter (ml). (Warrell 385). This level is readily obtainable in rats in the final stage of the disease. (Sebbane Fig2). The flea may thereby imbibe perhaps 5000 organisms. (Pierce 351).

Human-flea-human transmission is uncommon. For a human to transmit an infective dose to an uninfected flea the human must be highly bacteremic, i.e., have primary or secondary septicemia. The incidence of septicemia in humans is low compared to that in rats (Pollitzer7, 296), and even such humans typically don’t have more than 10,000 organisms per ml blood. (Pollitzer8, 65). A simple bubonic plague case cannot transmit the disease.

Secondary septicemia was probably more common in early modern Europe than it is today, when bubonic plague is treatable. But by the time it kicked in, the patient was probably isolated.

It is possible for an already infected flea to jump from one human to another, and thus infect more than one individual. However, that’s not likely to happen unless the first host is dead.

During the Third Pandemic, the Indian Plague Commission discovered that most bubonic plague patients did not have known contact with a prior case, there usually wasn’t more than one case per house, and if there were multiple cases in a house, they were usually simultaneous. If they were successive, it usually was associated with higher rat mortality. (Pollitzer9, 133). The implication is that the transmission was primarily rat-flea-human rather than human-flea-human. (134).

In the early 1900s, it was firmly believed that there was a serious risk of direct transmission of plague to humans from infected fleas on the persons or in the baggage of travelers, or in shipped goods such as grain, cotton, and hides. While that’s possible if the trip wasn’t too long, it’s more likely that such fleas would first cause rat plague. (Pollitzer7, 303).

If a human is bitten by an infected flea, the incubation period is typically 2–6 days. (CDC); there are rare reports of incubation for as long as 14 days. (Cipolla/Cristofano 121).

While dogs are resistant to plague, and probably not bacteremic enough to infect fleas, they may still carry infected fleas; sleeping in the same bed with a dog is reportedly a risk factor. (Weese 204). But being licked by an infected dog wasn’t. (Abbott 45).

By inhalation . Pneumonic plague may be transmitted by coughing or sneezing of infective droplets. CDC says that the incubation period for aerial transmission is typically 1–3 days. (CDC). However, Lien-Teh (1926) reported that it was 2–9 days with a norm of 2–6. (FDA 16, 29).

While pneumonic plague is contagious, it is possible to exaggerate the risk. First, “Person-to-person transmission requires face-to-face exposure within 2 meters of a coughing patient.” (WHO). Second, there are instances of people living in intimate contact with a pneumonic plague victim without getting sick. In the Manchurian epidemic (1921), there were four contacts living in the same 12×12 foot room as the first victim. Of these, only one contracted the disease. (FDA 17–18).

The last US case of human-to-human transmission of pneumonic plague was in 1925. (CDC). But from 1977–1998, there were 23 cases of cat-associated human plague in the Western US. Please note that this represented only 7.7% of the total 297 cases, so the cat vector is uncommon. They resulted in seventeen cases of bubonic plague, one of primary septicemic plague, and five of primary pneumonic plague. (Gage) The latter resulted from direct exposure to infective droplets.

Marmots are not only capable of direct (marmot-marmot, without fleas) transmission of pneumonic plague, in them the bacterium is pneumotropic—it heads straight for the lungs. In marmot regions of China, 45% of human contacts developed pneumonic plague. (Grent 196).

By ingestion . A human or animal may acquire plague by consuming an infected animal. The bacterium can survive for more than two months in a carcass. (Abbott 45). In the Middle East, humans have experienced pharyngeal plague by eating raw camel meat (20), and in the chaos of the Thirty Years War, it is not difficult to imagine a starving refugee or soldier consuming a rodent. Of 60 plague-infected cats, 75% had hunted. (Weese 204). That said, Simond showed that “it was actually rather difficult to convey plague to rats through feeding them.” (Orent 183).

Other routes. In the early-twentieth century, there was much concern that plague bacilli escape from the body in the blood, urine, and feces of septicemic patients (Simpson 213) but it now seems doubtful that these play an important practical role in transmission, unless these materials come in contact with an open wound.

Given what’s now known about the survival of the bacterium in the open, it’s likely that cases of apparent infection from contact with the clothing of a victim were the result of exposure to infected fleas living in that clothing, rather than the free bacterium. (Pollitzer2, 103). In other words, there’s no need to disinfect inanimate objects, only to de-flea them.

There have been reports of humans contracting plague after skinning a rodent, burying plague corpses, or in the case of children, holding “play” funerals for dead rats. (Abbott 29, 44–45; Chase 169; Treasury 147). However, it’s likely these too were from flea bites.

There are also rare reports of transmission by being bitten or kissed by a human pneumonic plague patient (Pollitzer9, 163), or bitten or scratched by a cat (Abbott 20) and I suspect that there was contact with infectious sputum.

Survivor Immunity

Those who survive an active infection (as opposed to being fortunate merely not to infected in the first place) probably acquire some resistance—and such individuals were once favored as attendants for later plague sufferers—but re-infection is known to have occurred. (Pollitzer3, 185). Still, the immune response is longer lasting (antibodies detected more than a decade later—Li) than in those receiving vaccines.

Grantville Literature

The literature available in Grantville concerning the plague is mainly limited to the encyclopedias, and what would be in the library of a rural general practitioner (Dr. Adams).

The 1911 Encyclopaedia Britannica (EB11) of course makes no reference to antibiotics, but these are identified in the “modern” (as of the RoF, April 2, 2000) EB.

Dr. Adams probably has basic medical microbiology, pathology, pharamacology and internal medicine books in his library but he probably never saw a case of plague before the RoF.

Don’t expect too much of his library; his medical microbiology textbook probably has just a portion of a chapter dedicated to Yersinia; see e.g. http://www.ncbi.nlm.nih.gov/books/NBK7798/#A1608.

Modern Anti-Plague Measures

Overview of Possible Protective Measures

With our modern knowledge of what causes the disease and how it is transmitted, we can logically classify protective measures as follows:

(1) diagnosing the presence of plague in humans or associated animals;

(2) killing rats or at least keeping them away from humans;

(3) killing fleas or at least keeping them away from humans;

(4) isolating humans with plague, especially pneumonic plague, from other humans;

(5) prophylaxis of exposed humans; and

(6) treatment of infected humans (and perhaps pets).

Improved Diagnosis

While swollen lymph nodes are a major indicator of bubonic plague, they can be associated with other diseases, including appendicitis, cellulitis (strep or staph infection), cat-scratch fever, Capnocytophaga canimorsus infection from dog or cat bites, sexually transmitted Lymphogranuloma venereum infection, infectious mononucleosis, cutaneous anthrax, measles, sleeping sickness, toxoplasmosis, certain snake bites, and tularemia. The location and painfulness of the affected node(s), the accompanying symptoms and the rapidity of their onset, and the circumstances (including geography, patient history, and other cases) must all be considered. (CDC; Gibson). However, it is prudent to assume that a febrile patient with buboes has the plague, until proven otherwise by more definitive tests.

Septicemic and pneumonic plague are less distinctive. Primary septicemic plague can be confused with septicemia caused by other bacterial infections. Primary pneumonic plague can be confused with pneumonia caused by other pathogens. The initial differentiation is typically on the basis of plague being more fulminant (rapid and intense) and of plague outbreaks being more explosive. (CIDRAP).


In the twentieth century, plague was confirmed by culturing a blood or node sample (not without risk!) and studying any bacteria observed under the microscope. Yersinia pestis is a slow grower; it takes about 48 hours for the colonies to reach 1–2 mm diameter. Growth is faster at 28oC, but incubation at 37oC is needed for expression of the diagnostic F1 antigen. (CDC/BLP).

Perhaps because of its finicky growth characteristics, there have been several instances in the early-twentieth century of cases with negative bacteriological findings that were belatedly recognized as plague. (Simpson 375ff).

Microscopic identification of the bacteria was facilitated by use of various specific stains. The Gram staining procedure (crystal violet, iodine, alcohol or acetone, safranin) leaves gram-positive bacteria purple and gram-negative ones pink; Yersinia pestis bacteria are gram-negative rods.

But many other bacteria are, too. The Wayson stain comprises basic fuchsin, methylene blue, ethanol, phenol and water. The Giemsa stain is methylene blue, eosin and Azure B. The original (1890s) Wright’s stain was an alcoholic solution of methylene blue and eosin Y; modern formulations are likely to also contain azure A and thionin. Stained with any of these reagents, Y. pestis often has a bipolar “safety pin” structure. Biochemical tests may also be helpful; Y. pestis is oxidase, urea and indole negative, and catalase positive.

During the San Francisco epidemic, a simple agglutination test was introduced. Blood from a suspect was mixed with the Yersin antiserum (see below) and clumping indicated that the bacteria in the blood were Y. pestis . (Chase 97). In Surat 1994, antibody tests were positive even though bacteriological tests were negative. (Marriott 159).

By 1996 CDC guidelines, plague is reported as “suspected” until supporting lab reports are available. Detection of F1 antigen in a clinical specimen, or of an elevated serum antibody titer to F1 antigen, is considered presumptive of plague. If the elevation is four-fold or greater, or the bacterium is actually isolated from a clinical specification, that’s confirmatory. (More recently, phage lysis and DNA hybridization tests have become available.)

While the resident Grantville doctor (Adams) is likely to have books or articles mentioning the F1 test, they are likely to be from a clinician’s perspective, and might not even identify F1 as a capsular antigen, let alone provide any specifics as to how to purify or identify it.

Even with modern technology, diagnosis of plague can be belated. There was a case of a community hospital in California failing to recognize primary septicemic plague until day 9. (Margolis).

While not strict proof that a suspect human illness is plague, human plague is more likely if rats in the vicinity are found to be infected. (Pollitzer5, 317).

Identification of Plague Sites

A new victim may have acquired the disease at home, at work, or otherwise (restaurant, tavern, theater, etc.). It’s important to determine as quickly as possible which location was the source of the disease, because plague is a “disease of place” (more precisely, of infestation with infected fleas). This can be done, for example, by inspecting the premises for signs of rat infestation and also attempting to trap rats and fleas. (Creel 1891). A possible complication is that a victim may have been moved from the place of infection to fool the authorities.

Improved Isolation

Bubonic and septicemic plague are not themselves contagious. In the premodern period, a practical advantage to isolation of non-pneumonic cases was that even though they weren’t contagious in the strict sense, they were probably flea-ridden, and if they weren’t, they offered contagious blood meals to any fleas that happened upon them.

However, modern practice would be to treat the patient and “contacts,” their clothing, and the suspect premises with insecticide and also to eradicate resident rats, and rat-proof the building, so infection would not recur. Thus, the modern justification for isolation of those suffering from those ailments is that they could, at any time, develop secondary pneumonic plague—which is contagious. In any event, isolation need not be longer than the length of time that the patients remain in an infective state.

For bubonic plague cases, isolation is likely to be for two to three days after commencing antibiotic treatment, and the hospital might additionally insist on it continuing until clinical improvement (e.g., fever reduction) is seen.

Those with confirmed pneumonic plague remain in isolation under respiratory droplet precautions (caregivers wear gowns, gloves, eye protection, and surgical masks) until all lab cultures are negative. In addition, close contacts that refuse prophylaxis are quarantined for seven days. Isolation, by the way, is strict; in Surat 1994, the militia had orders to shoot-to-kill escaping patients, (Marriott 51); the escapees were called “living bombs.” (106).

Improved Quarantine

The shutting in of whole households is a bad idea. It’s important to separate those who are isolated because they actually have the disease from those who are quarantined because they are “contacts” or otherwise “suspect.” Otherwise, the latter are put at greater risk. Such separation proved productive in 1900 Honolulu. (Mohr 162).

As noted earlier, even if the victim is not strictly contagious, it’s likely that infected fleas are present and if given the opportunity, will transmit the disease to other members of the household. We thus must either remove the contacts to a safe location (while making sure that the fleas are not transported with them) or eliminate the fleas on the spot. In general, the latter is preferred (forced relocation is quite unpopular), but if the household is heavily infested, a temporary relocation may be in order. (Pollitzer 530).

We now know that the traditional quarantine period was far longer than necessary; it greatly exceeds the bacterium’s incubation period. Of course, the incubation period itself isn’t fixed; there’s a range, and one may be conservative and pick a quarantine period that exceeds the longest reported an incubation period, or a somewhat shorter one that has less impact on commerce (and is less likely to be flouted by those who find quarantine oppressive).

EB11/Plague refers to the Venice convention of 1897, under which ships were considered infected if there had been a fresh case within the last twelve days. More particularly, it classified ships as being healthy (left infected port at least ten days before yet had no cases of plague on board), suspect (had cases, but not within the last twelve days) or infected. Crew and passengers from healthy ships were just placed under surveillance for ten days (they could go about their business, but doctors checked up on them). Those from suspect ships were treated the same way, but the ship was disinfected. Only those from infected ships were quarantined, for up to ten days. (Simpson 355). The 1903 convention reduced the quarantine to five days and required that all rats on an infected vessel be destroyed. (358).

A further problem with down-time practice is that the quarantined persons were in contact with each other. Imagine that someone who was on the ninth day of quarantine (thus, unlikely to be infected) is brought into contact with a new arrival. If the new arrival is infected, and is coughing, or there are fleas present, the bacterium could be transferred to the earlier suspect.

It’s probably too much to expect “suspects” to be quarantined in individual partitions but one can at least provide a multiroom facility such that all “suspects” in a particular room entered quarantine on the same day.

As in early modern Europe, a cordon sanitaire around a plague district proved beneficial as a stopgap in circa 1900 Honolulu and San Francisco. However, since such districts are likely to have concentrations of the poor, and thus perhaps of particular ethnic, racial or religious groups, it may be difficult to say where good public health practice ends and discrimination begins. It was not unusual for quarantine borders to skirt buildings owned by privileged citizens, or for the border to be more porous where they were concerned. (Chase 68). Also, con men claimed to be able to raise the restriction for a suitable price. (69).

Also as in the past, the use of certificates of health to control movement encouraged the desperate to obtain clean certificates by bribery or forgery (68), and the appointment of citizens to monitor their fellows’ health resulted in abuse. In 1900 Honolulu, both members of the Citizens’ Sanitary Commission and quarantine camp guards were accused of theft, sexual harassment and attempted rape. (Mohr 115ff).


The bacterial cause of plague was discovered in 1894, but the role of fleas was not accepted by the scientific community until 1907. In the Third Pandemic, a variety of chemical disinfectants were used. For example, EB11/Plague suggests disinfection of victims’ premises with “corrosive sublimate” (mercuric chloride) or sulphuric acid. Carbolic acid and chlorinated lime were used during the SF epidemic. (Chase 125). However, it is likely that these disinfectants’ purpose was to kill “free” bacteria, and any effect on infected fleas was purely serendipitous. In 1914, Creel (1901–2) declared the use of germicides “a waste of time.”

It’s difficult to predict whether some down-timers will take the advice of EB11 even though it’s dubious given what more modern medical books say about plague transmission.

Condemnation and Destruction of Buildings

If plague cases continued to occur in a building that had been “disinfected,” the city might order the building condemned and destroyed. In San Francisco’s Chinatown, porch and balcony additions were often condemned as unsanitary (and indeed they probably offered harborage to rats) and demolished. Unfortunately, the scrap wood was valuable as firewood and the indigent collected the wood for sale or personal use and subsequently contracted plague. Hence, the authorities had the debris treated with powdered lime. (Chase 126ff).

In 1899 Honolulu, this was taken a step further; if a plague case arose in a wooden home, the household would be relocated to quarantine quarters, their belongings would be disinfected if possible, and burnt otherwise, and the home (and possibly also adjacent wooden structures) burnt. (Mohr 88, 90). Places of business were handled similarly. (176, 201).

As plague cases increase, there will be demand to define a “plague site” more broadly, and thus to incinerate blocks or even districts instead of individual buildings. (Mohr 122).

But incineration is a dangerous solution, as the conditions that make a district favorable to rats also are unfavorable to controlled incineration. In Honolulu, the Chinese consul warned that in Chinatown, the buildings stood so close together that it would be difficult to burn them separately, and he advised tearing them down first and just burning the rubble. (Mohr 149).

Nonetheless, on Jan. 20, 1900, Honolulu firemen began a controlled burn of “block 15.” The wind shifted and they lost control of the blaze. The result was the loss of one-fifth of the city’s buildings, and the displacement of one-eighth of the population. (Mohr 4, 125ff).

Despite this tragedy, burning continued to be a major tool in the plague-fighting arsenal; Kahului’s small Chinatown was incinerated. (Mohr 172). International observers were impressed by the Hawaiian success, and fire was used at Kobe 1901, Mazatlan 1903, and Manchuria 1910. (199ff).

There was also the issue of what to do with the plague site. In Bombay, the British “plowed the ground . . . , soaked the soil with petroleum, burned them, and kept the site isolated for a year.” In Honolulu, soil samples were taken and, if the site were free of the plague bacterium, it could be rebuilt. (186).

Incineration, of course, increased the cost to the affected household, and this as well as fear of relocation in turn increased reluctance to report suspicious illnesses to the authorities, and indeed led to bodies being found in “improbable locations.” (162). Since the fires were deliberately set by the city, the insurance companies refused coverage; all claims had to be directed to the government. (182). Recovery was typically half the face value of the claim, even with federal assistance. (191).

It’s uncertain whether incineration by itself had any effect on rats (and fleas) in burrows beneath the burnt building, and even above-ground rats were probably scattered rather than killed unless the site was first surrounded by a guarded “stockade.” To deal with the subterranean vermin, the site would have to be dug up after burning and treated with rodenticide and insecticide. (Creel 1903).

Better control methods were developed later in the twentieth century and by 1952, burning down plague-infected houses was considered a “barbaric practice.” (Pollitzer 529).

Rat and Flea Control

It is important to prevent, as much as possible, the exposure of humans to flea ridden rats in the first place.

However, public health experts have changed their mind as to how to proceed. In the early 1900s, Bishopp taught that while “the elimination of all fleas is desirable,” “the prime move should be against the rats which act as hosts for both the plague bacillus and the fleas that carry it.” This move, he thought, should involve both rat killing and rat-proofing. He recognized that this would involve exposure to the fleas, but suggested that the workers could take precautions against flea bites. (Pierce 367–8). In 1907, US Public Health surgeon Rupert Blue declared, “if we destroy the host [the rat] there is no longer danger of infecting the parasite [the flea].” (Treasury 146).

Modern thinking is that it is a grave error to attempt to kill the rats without previously or simultaneously killing their fleas. Fleas will leave a dying rat and seek out a new host, which might be a human or a pet. (Webber; Acha 215; Beran 108; Abbott 47). Lisbon reported that in the absence of plague, only one of 246 fleas caught on men was a rat flea, but of the 30 fleas found on men in a boarding-house during a plague, 14 were rat fleas. (McFarland 592).

Considering how much controversy there was over this issue in the twentieth century, I suspect that some down-timers, even if they accept the Grantville teaching of the rat-flea connection, will insist on dealing with the rats first. Perhaps making an outbreak worse rather than better.

Rat Monitoring

In early-twentieth century San Francisco, dead rats were routinely examined for signs of plague. A single trained bacteriologist with proper facilities can examine up to 100–150 rats/day. (Simpson 363). The corpse was immersed in a killing solution to kill off fleas and then dissected. Plague rats often exhibit buboes, and another indication is the appearance of the subcutaneous blood vessels. (Treasury 32ff). This monitoring was important, as in Manila it was found that when the level of infection reached 2%, transmission to humans reached epidemic proportions. (Chase 168).

It was important to keep track of where the rats were found. In Hong Kong 1894, it was recognized that “the presence of plague-infected rats in a house or locality meant, sooner or later, if immediate measures of precaution were not taken, cases of plague.” (Simpson 216).

Care must be taken, in collecting dead rats, not to contract plague. In 1901 Hong Kong, 30% of the collectors died of the disease. (217). The key is flea control. In Madagascar 1995, a three-chamber trap was used. The rodenticide-poisoned bait was in the central chamber, and the end chambers, through which the rat had to pass, were charged with insecticide. Thus, the fleas were killed before the rat. (Marriott 182).

Reducing Flea Transmission to Humans

To bite you, the flea must reach you. There are basically two approaches to preventing this; interpose a barrier between the fleas and our bodies, or trap or kill the fleas beforehand by mechanical or chemical means.

A flea can jump up to 13 inches horizontally and 8 inches vertically. (Pierce 369; Pollitzer7, 302). We are stationary targets when we are asleep, so it’s important to secure beds from attack. The bed should be elevated more than eight inches off the floor, and the bedding should not hang down. The legs may be placed in pans of water, and ideally these are covered with a film of kerosene. Flypaper may also be placed on floor around the bed, in strips at least thirteen inches wide. (Bishopp/FB683 13). Of course, this presupposes that the fleas aren’t already living in the bedding or on the sleeper.

Fleas may be manually removed from a person or animal with a flea comb. The fine-toothed comb for removing head lice dates back to the Neolithic. (liceworld.com). This is definitely down-time tech, my only concern is that the best time to remove fleas is before they get infected with plague.

Mechanical Trapping and Killing of Fleas

A popular flea trap involves a candle at the center of a bowl of soap water. The fleas are attracted to the heat; they jump and end up in the water; the soap is to reduce the surface tension and help trap them there. (Klein 43; Bishopp 29). Instead of water, one can use paper covered with a sticky substance.

In February 1994, Popular Science (46) described a flea trap that used an intermittent green-yellow light as a lure; the theory was that the flies would think that the interruption of the light was because a person was passing by and would jump at the presumed shadow. I suppose that the effect could be substantially duplicated by down-timers with a candle, colored glass and a wind-up clockwork shutter, but I doubt the incremental advantage would warrant the cost.

Both humans and animals have been used as portable flea traps. In the early-twentieth century, a Cornell University building “was cleared of fleas by having the janitor, with legs wrapped in fly paper, walk back and forth in the infested rooms.” (Bishopp 28). The Indian Plague Commission used guinea pigs as flea collectors (29), although I think the motivation was sampling rather than control. In San Francisco, they were used by sanitation workers to reduce the flea population before they entered a plague house. (Chase 183). These “furry flea magnets” would be even more useful if they were treated with insecticide first.

Flea traps are effective in detecting the presence of fleas, but their effectiveness for control has been disputed (Vail). I suspect that they are more useful for disposing of “scouts” than for eradicating an established infestation.

It has long been recognized that vacuum cleaners collect fleas (Heloise 129). Vacuuming removed 15–27% of larvae and 32–59% of eggs introduced into a carpet. Pupae weren’t affected, as their cocoons were bound to the carpet. (Byron). Recently, it was discovered that vacuumed fleas are actually killed, possibly because the vacuum brushes wear away the cuticle (which keeps the fleas from drying up). (Hink).

The first vacuum cleaners (1860s) used a bellows to create the vacuum. Ideally, pushing the vacuum cleaner across the floor would cause the bellows to open and close. (In the McGaffey “Whirlwind,” you had to turn a crank as you pushed.) A gasoline-powered vacuum cleaner was patented in 1899 and an electric vacuum cleaner in 1908.

Food-grade (natural) diatomaceous earth can be used to kill fleas; the silica shards degrade the flea cuticles. However it’s effective only when dry. (Kern 3).

If just adult fleas were present, then storing contaminated goods in a insect-tight box for a week would be sufficient for them to starve and die. But if there are eggs or larvae present, you’re out of luck, as they can dine on the textile fibers.

Flea Repellents and Insecticides

Strictly speaking, a repellent repels but doesn’t necessarily kill, and an insecticide kills, but doesn’t necessarily repel.

These compounds may be applied to people’s skin, clothing, and bedding, and to the fur of pets such as cats and dogs.

Rats may be used to deliver insecticides to places where fleas linger that humans can’t access directly. Thus, if rat runs are discovered, these are dusted with insecticide (that isn’t also a rodenticide). The rats pick it up on their fur, and carry it into their nests. If the access tunnel to a rat burrow is found, then the insecticide can be blown in. (Webber 252).

The modern EB/flea indicates that adult fleas on infested animals may be killed with pyrethrum, rotenone, malathion and methoxychlor, and flea larvae and pupae, away from the host, with dieldrin, lindane and malathion. Benzyl benzoate, phthalate compounds and diethyltoluamide are identified as repellents. The essay doesn’t mention DDT, no doubt because its use was restricted in 1972. There are many cat and dog owners in Grantville, and they may own samples of additional insecticides.

DDT (dichloryl diphenyl trichloroethane) is both insecticide and repellent. NTL, the USE delegation to England brought with it “several pounds of the DDT which the fledgling American chemical industry was starting to produce.” (Flint, 1633 , Chap. 21), and it was used specifically as a preventative against plague. There aren’t a lot of date cues as to when the delegation left home (Eric’s motto is “vague is good”); John Bogan’s proposed timeline has the delegation departing in June, 1633.

Historically, the first large-scale use of DDT was in WW II, to control lice-borne typhus. A powder with 10% DDT was applied to “to the patient, his clothing and bedding.” Over 1,300,00 Neapolitans were treated in the first month and civilian cases were halved after the first week. (Bate).

Ultimately, DDT was delisted because of its environmental effects (documented in Carson’s Silent Spring), but it is still used occasionally under emergency exemptions: “In January 1979, DDT was used to suppress flea vectors of murine typhus in Louisiana.[26] As late as June 1979, the California Department of Health Services was permitted to use DDT to suppress flea vectors of bubonic plague.[27]” (Bate).

DDT can be sprayed as either a dry powder or as a solution, and wherever “a DDT residue will be left in places where [the fleas] develop, feed, seek shelter or will crawl over the deposit in the course of their normal activities.” The residue is effective for several months, sometimes a year or more. (Henderson).

Malathion (2-(dimethoxyphosphinothioylthio) butanedioic acid diethyl ester) is presently the most commonly used insecticide in the United States, and it does kill fleas.

Carbaryl (1-naphthyl methylcarbamate) was first used in 1976 and is the third most used insecticide in the United States.

DEET (N,N-diethyl-meta toluamide) (CCD 299) was first synthesized in 1954. (Eldridge 572). It’s just an insect repellent, activating an olfactory receptor.

Icaridin (Picaridin) was not used as an insect repellent until 2001, which was after RoF.

Benzyl benzoate was adopted at the end of WW II as a standard clothing repellent, as it was shown to be effective through at least two soap-and-water washings. It’s the simplest chemical of the repellents known to me, and over 90% effective against fleas, but it was “poorly accepted by military personnel” because of its disagreeable odor and tendency to irritate the skin. (GPO 511).

Hexachlorophene (Lindane) is an insecticide and antiseptic, and kills more quickly than DDT, but its period of effectiveness is shorter. (Pollitzer 522). It’s quite neurotoxic and also causes contact dermatitis. (EPA 204). I would avoid it.

I have documented the existence in chemistry books reasonably believed to be in Grantville (Condensed Chemical Dictionary, Merck Index, Morrison & Boyd’s Organic Chemistry) of guidance for the synthesis of DDT, malathion, carbaryl, DEET, benzyl benzoate and hexachlorophene.

Pyrethrins are natural insecticides of botanical origin. The species producing the most potent insecticide is the Dalmatian chrysanthemum (Tanacetum cinerariifolium , syn. C. cinerariifolium ) and the big question is, would it have been in Grantville, in a garden or indoors, at the time of the RoF?

If not, we would need to find it in nature. Matsuo (3) says that it is native to Dalmatia, that the insecticidal properties of its flowerhead extract (pyrethrum) were known locally as early as 1694, and that these properties were scientifically verified around 1840. But according to Casida (3), it was a variety developed in Dalmatia in the nineteenth century. Neither Matuo nor Casida cite sources. The first scientific description of the species was in 1847. (Grdisa)

I am inclined to believe that it was indeed native to Dalmatia, since Dalmatia was hardly a plant breeding center. But that still begs the question of what was its native host range in the 1630s. Once the pyrethrum powder became a commercial product (1850s), it would have been planted in places it didn’t grow originally. However, it presently grows in the wild in “in southern parts of Istria (Premantura), Kvarner islands (Krk, Cres, Lošinj), mountains Velebit and Biokovo, and along Dalmatian coastal region and its islands ( Brač, Korčula, Dugi Otok, Mljet, Vis, Hvar, Pag and Pašman) (Grdisa).

Within Europe, it can also be grown in France, Italy, Spain and Switzerland. The first plantings in Japan were reportedly of German origin. (Matsuo). “Commercial cultivation of the Dalmatian Chrysanthemum is done at altitudes between 1600 and 3000 meters because pyrethrin concentrations increase with altitude, whereas flower production diminishes at altitudes below 1600 meters. Ideal growing conditions are a semi-arid climate with cool winters, ca. 1200 mm rainfall and a two- to three-month dry season. They tolerate temperatures as low as -12oC.” (EAP).

In the early-twentieth century, yields were perhaps 500–600 pounds dried flowers per acre. (Wardle). With modern cultivars, “Around 250 kg of dried flowers are produced per hectare during the first year, increasing to 1000–1200 kg per hectare in the second and third year. After third year yield declines.” (Grdisa).

The active agents are cinerin I and II, and pyrethrin I and II; the structures appear in the Merck Index, but without synthetic guidance. They are rapidly inactivated by air.

Other chrysanthemums used to make insecticidal extracts include T. marshalli , carneum, coccineum (roseum ), and coronarium , the latter two allegedly used in Persia in ancient times (Panda 1), but their levels of the active agents are much lower than for the Dalmatian species. One or more of these were used in the “Persian insect powder” sold in the early-nineteenth century, but replaced by the Dalmatian extract. EB11/Pyrethrum says that roseum is found in the Caucasus and blooms in May and June.

Pyrethroids are synthetic analogues of the natural pyrethrins. The goal has been to increase potency, selectivity and stability. I think they are too complicated to be reasonable synthetic targets for the NTL 1630s.

Rotenone is a fairly complex chemical that was isolated in 1895 from a plant (Lonchocarpus utilis) of French Guiana—the natives used a crude extract as a fish poison. In canon, there is a USE colony in Suriname (Dutch Guiana) and there’s a reasonably good chance that a similar plant is available there.

Sodium fluoride is able to kill flea larvae (Pollitzer 525) and may be useful in combination with agents that specifically target the adults.

Methoxychlor is a symmetric aromatic compound that is no harder to synthesize than DDT and indeed was introduced in 1948 as a DDT substitute. The “modern” EB portrayed it as “greener” than DDT but EPA banned it in 2002.

Chlordane is another organochlorine, and has proven effective against DDT-resistant fleas (Pollitzer 523), but it’s probably too complex chemically to be synthesized within the first decade NTL and it was banned by EPA in 1988.

Sulfur is already available in the 1630s, and it can be used to repel some insects. Unfortunately, the rotten egg smell (from oxidation) is pretty good at repelling people too. Moreover, despite folk claims on its behalf, it’s apparently ineffective for flea control. (Miller 286; Halliwell). But don’t confuse dry sulfur with the fumigant sulfur dioxide!

Insect growth regulators (lufenuron, methoprene, pyriproxyfen) only affect immature fleas and are likely to be too complex chemically to be available in the NTL 1630s.

Boric acid is available from the Maremma of Italy. Boric acid will kill flea larvae that ingest it, but has no effect against adult fleas. (Vail).

Kerosene-water emulsions were used in early-twentieth century China to kill fleas. (Treasury 233), and in 1914 Creel (1901) suggested using them to wash the walls, floors and ceilings after a building was fumigated. Vedder (108) suggests sprinkling a floor with kerosene, and smearing posts with crude oil, to kill the flea larvae. Note that kerosene can be made from coal tar, and coal is available in England, France and Germany.

Flea repellent claims have been made for various botanical products (bayberry, pennyroyal, eucalyptus, rosemary, citronella) as leaves or oil extracts, without any documentation of effectiveness. (Vail)

Creel argued in 1910 that “measures intended for the destruction of fleas are . . . of relatively small value,” urging that rat-proofing is more effective as an anti-plague measure. (Treasury 170). I disagree . . . .

Rat Control

Rat control—whether in the form of rat-proofing or extermination—is complicated by seventeenth-century building methods. London featured many “large blocks of interconnected houses built from the debris of older buildings.” (Hall 64). These can be processed effectively only on a block by block basis. Even in the case of new construction, the buildings often shared side walls. (67).

In Manila 1905, the authorities began rat control by mapping the sites at which plague-infested rats had been found. They then fanned out from there, establishing the furthest points away at which such rats could still be found. The rat-catching force was concentrated along the borders of these plague rat regions, and moved inward. They were followed by the rat-proofers. Once all regions were processed, the city was divided into a grid and rats were caught for testing each week at each point on the grid, so any new occurrences could be detected and controlled. By 1906, there was no longer plague among human beings, and by 1907, it had also been eliminated from the rats. (Treasury 205ff).

Since the rats might be infected, they must be handled with care (and insecticide). In 1900 Honolulu, the winner of a rat-catching contest came down with plague. (Mohr 164). Subsequently, the public health officers warned citizens to use shovels to dispose of their trophies. (Mohr 167).


There would be good public health and economic reasons to control rat populations even if they didn’t act as plague vectors. Hence we want to make our towns less desirable places for them to live and breed, by reducing rodent habitat (hiding and nesting places) and barricading them from potential food supplies and human workplaces and habitations.

This isn’t easy. They can squeeze through “a hole no larger than a quarter dollar,” jump upward two feet (three with a running start) and four feet horizontally (more if jumping from a height), and climb a vertical wall if they can find a claw hold or brace their backs (say, at an inside corner). They can survive a five-story fall and tread water for three days. (Hendrickson 87).

Where do we start? Considering how the black rats spread from place to place, one possibility is the waterfront. That means rat-proofing the ships (by fumigation) and the wharfs (by rat-resistant construction). (Treasury 177).

Ships . A ship can harbor an amazing number of rats. In San Francisco, a 260-ton lumber-carrier was fumigated, and 310 dead rats were collected. A grain-carrying vessel yielded 1700 rats. (208).

We also must address the avenues by which rats get off ships. Early-twentieth century authorities required that ships be fended off six feet from the dock, gang planks raised at night, and mooring lines equipped with “rat funnels.” Hurdy recommended that these funnels be of heavy galvanized iron, with a 3 inch diameter spout and increasing to at least 36 inches diameter at the other end.

Of course, the rats may get a free ride to shore if they are hiding in the cargo. Fumigating vessels with cargo is more complicated than fumigating an empty ship, as you must avoid damaging the cargo or starting a fire. (Treasury 212ff). There is the option of transporting the cargo to an off-ship fumigation chamber.

Rail cars . Rail cars and freight yards likewise can offer food and shelter to rats.

Buildings . Rat-proofing buildings isn’t easy; one must worry about the floor, walls, ceiling, doors, windows, air or light shafts, water and sewage pipes, and wiring entry points. The rats only need to find one point of ingress.

The most important buildings to rat-proof are those that contain large amounts of food. Creel says, “the food depots . . . in order of importance are stables, meat markets, bakeries, restaurants, groceries, warehouses and private dwellings.” (Treasury 174). However, a building should be given higher priority if it is known to be a plague source. And then, having processed that building, attention should be given to other buildings on the same block.

In terms of rat-proofing buildings, there are two basic approaches. During the San Francisco outbreak, a Missouri farmers’ method of keeping rats out of corn bins was used to protect the cottages at a quarantine camp; the cottages were “mounted on stilts eighteen inches off the ground.” (Chase 183). This is called “rat-proofing by elevation”; the eighteen-inch clearance provides room for their natural enemies (cats and dogs) to hunt them.

This is a pre-RoF technique. The Norwegian Folk Museum in Oslo has a circa 1300 “loft” storehouse from Telemark (building 133). This storehouse was mounted on stilts of a special design. On top of each wood column, a split-log is placed, turned so the flat side was down—the combination looks in profile like a mushroom. The goal was to create an overhang that prevented rats from climbing up the column into the building.

The second approach is to use walls and flooring that are made of a material the rats can’t gnaw through quickly, e.g., concrete or stone. (Treasury 161, 176). Note that a rat can chew through four-inch-thick concrete and half-inch-thick metal (rat-patrol.org). It may be helpful to embed a wire mesh in the concrete.

I fear that this structural rat-proofing will be very expensive in the NTL 1630s. Stone ice houses, for example, were much more expensive than the wood ones introduced in the nineteenth century. Concrete, in the OTL 1630s, was a forgotten material (despite the Romans), and it will take time for concrete technology to diffuse out of Grantville. Steel, while known, was even more expensive than stone.

Unfortunately, the places where the rats are most likely to congregate are the neighborhoods that can least afford sanitation measures. In Newcastle, the plague outbreak occurred first in Sandhill, the poor quayside area.

In twentieth-century America, rat-proofing was required by municipal ordinance, i.e., the cost was imposed on the property owner. If that’s done in the NTL 1630s, there will probably be a lot of resistance. Public perception of success will help overcome that resistance, but bear in mind that substantial results won’t be seen until rat-proofing is the norm. There may be an advantage for the central government to offer financial incentives to the first towns (preferably smaller ones) to rat-proof, and progressively reduce the incentives as the proof of success mounts. Ultimately, of course, the towns will punish those who fail to rat-proof, rather than pay those who do.

Creel (1896) suggests that rat-proofing measures be urged “in the early days of an epidemic, when every suggestion of the sanitary authorities is eagerly supported by the community.” But it would be better to introduce them before the first case of plague.

Some rat-proofing measures fall short of wholesale rebuilding of structures. In stables, one may require metal-lined feed bins. For residential kitchens, one can similarly provide rat-proof containers. Sanitary garbage cans (metal with tight covers) would be helpful, and of course the garbage must be removed frequently. (Treasury).

Rat Extermination

Rats may be exterminated using rodenticides, rat pathogens, traps, or natural predators. Professional rat catchers existed before the RoF (the first appearance of the name “Pied Piper” was in a book published in 1605), and they employed all three methods. Less directly, one should cut off their food supply, destroy their existing nests, and prevent new nests from being established.

Extermination, by itself, is not an effective control technique. When a substantial number of rats are killed, the survivors become more wary of traps and hunters, and also they face less competition for the available food. The decrease in rat population is temporary. (Creel 1896).

Rodenticides. Fumigants kill by inhalation. Since they are often effective against fleas as well as rats, we discuss them in a separate section. Here, we consider rodenticides that kill by ingestion or skin contact.

Chronic (slow-killing) poisons have the advantage that the survivors are less likely to learn to avoid the bait. The bait is first left without the poison and once it is taken the poison is added. The preferred chronic poisons are anticoagulants like warfarin, bromadiolone, and diphenadione. Unfortunately, their chemical structures are complicated enough that I doubt that they will come into play in the NTL 1630.

A pioneer method that supposedly killed rats was to feed them flour laced with plaster of Paris; this supposedly formed “enteroliths” that killed them (Creel 1894). This bit of folklore was tested in 1990 and no rats died after two weeks of this diet. (Meade).

So that leaves the acute poisons. The poisons used in the early-twentieth century were arsenic, white phosphorus (OTL 1667)and strychnine (it comes from the seeds of an Indian tree, and reached Europe in OTL late 1700s) (Creel 1894). Arsenic (trioxide), sold as “ratsbane” in the 1500s, is relatively cheap and slow acting (so the poisoned rats die in their burrows, away from humans); Pollitzer prefers it to phosphorus. Strychnine is only poorly accepted by rats. (Pollitzer 463ff).

The bulbs of an onion-like Mediterranean plant (red squill, Urginea maritima ) contain scilliroside, a potent rodenticide, and extracts have been used against mice at least as early as the thirteenth century. It is one of the safest rat poisons because it causes humans and pets to vomit (rats can’t vomit!). The principal problem is batch-to-batch variation in potency, so each batch should be tested for adequate potency (say, LD50 of 500–600 mg/kg vs. rats). (Pollitzer 407ff; Verbiscar).

The metal phosphides are probably the most popular nowadays. When digested, they generate phosphine gas. Zinc phosphide will probably be the easiest one to produce and it’s more potent then either arsenic or red squill extract.

Rat pathogens . EB11/Plague makes reference to the Danysz virus but provides no particulars as to where it came from or how effective it was. It turns out that it is not actually a virus at all, but rather a bacterium. I have seen it identified as Bacillus typhimurium (Chase 136) or asBacillus enteritidis (Besson 444). While the San Franciscan public health officers thought it effective (Chase 136), Rosenau determined that it is only fatal to rats if eaten in large amounts; in smaller doses it renders the rat immune. (Treasury 181ff). Moreover, evidence has mounted that it causes gastroenteritis in humans. (Pollitzer 496).

Traps . Human ingenuity has come up with a variety of traps, baited or otherwise. Traps may imprison, immobilize or kill the rodent. The imprisoning (cage) traps may either drop down on the vermin or permit entry through a trap door that shuts when triggered. The immobilizing traps may have glue boards or steel jaws. The killing traps may have a spring-loaded wire (snap) to break the animal’s back, or they may be barrel traps with tipping tops to drown it. (Treasury 154ff; Pollitzer 460). Rats are clever, and will learn to ignore traps, or to snatch the bait without getting caught. Estimates of the number of traps one person can handle vary from 70 (Creel 1893) to 200. (Pollitzer 462).

Natural Predators . Webber says that “a well-trained cat can be most efficient”—I will ignore the question of precisely how one trains a cat—but some older sources are not fond of felines. Creel (1895) declares they are all of “doubtful value” and can carry infected fleas to their owners. Rucker warns that the ordinary cat won’t attack a large rat, and favors dogs, especially terriers, as ratters. (Treasury 159). But bear in mind that Rucker was writing in 1910, when the large brown rat occupied the city. In the seventeenth century, the cats have the opportunity to pursue the black rat, a smaller prey. (167).

In the countryside, there are lots of predators: weasels, ferrets, minks, badgers, skunks, hawks, and owls. However, farmers don’t give them much encouragement, and in fact we’re lucky if there aren’t bounties for killing them.

Professional urban ratcatchers have been known to use ferrets, and sometimes they are trained to work with dogs. (Treasury 168). A Middlesex ratcatcher with two ferrets killed 250 rats in one barn in a day’s work. (Forbush 13).

Incentivization . In some places and times, bounties have been paid for each rat killed. Human nature being what it is, people abused the bounty system. A 1907 Danish law punished “any person who preserves or breeds rats, or imports rats from abroad, in order to obtain premiums . . . .”


Fumigation has the advantage that it can combat rats and fleas simultaneously. A fumigant is a substance which, under expected temperature and pressure conditions, can be produced in the form of a gas at a fatal concentration for a target harmful organism. Fumigants kill by inhalation; at least partially by asphyxiation (displacement of oxygen), and in most cases by a direct toxic mechanism, too. Thus, DDT, which requires direct physical contact between the flea and the DDT residue to act, is not a fumigant.

Generally speaking, fumigants are not insect-specific, they kill mammals, too. That means they will kill rats. It also means that they will kill unwary humans.

In terms of use, it’s important to recognize whether the gas is flammable (F) or non-flammable (NF), and whether it’s heavier (HTA) or lighter (LTA) than air. HTA gases sink and diffuse slowly; LTA gases rise and diffuse rapidly. Fumigants which boil at above room temperature (HB) have the advantage that they can be handled as liquids, whereas low boiling (LB) fumigants must be stored and transported under pressure and/or refrigeration.

With regard to all of the fumigants described below, Grantville literature is unlikely to go beyond identifying the chemical as a fumigant; effective concentrations/durations will have to be determined by experiment.

Sulfur dioxide (NF, HTA, boils -10°C) is the earliest known fumigant, obtained by burning sulfur. Odysseus orders, “Bring sulfur, old nurse, that cleanses all pollution, and bring me fire, that I may purify the house with sulfur . . . ” Sulfur was specifically used as a plague disinfectant in 1720 Marseilles (Byrne 121), without of course any concept that it might help by killing rats or fleas.

In 1914, it was virtually the sole ship fumigant in use at American ports and even in 1932 it was used 6.5% of the time in the continental US and 89% in the Philippines. (Williams). Even after hydrogen cyanide displaced it at major ports, it remained in use at those with occasional needs for fumigation, because the fumigation crews didn’t require the same stringency of training. Sulfur dioxide has the advantage that it can be smelled at concentrations much lower than those that are dangerous to humans. (Pollitzer 505).

Sulfur dioxide will kill plenty of rats, but the smell alerts them to danger, and there are many places for them to hide where air circulation is poor—in coal bunkers, dunnage, pipes and bilges, and behind sheathing. The steamship Innamincka was fumigated five times over a two-week period, each time killing more rats. (Grubbs 1267).

If water (perhaps as steam) is present, sulfur dioxide will react to form sulfurous acid, which is a parasiticide and germicide. (Wardle 156).

There are several different methods of generating sulfur dioxide. (Rosenau 1139ff). The simplest is to burn sulfur inside an iron pot at the fumigation site. A chemist can calculate the theoretical amount needed for complete combustion in air, but the practical rule is that one pound produces 1% gas when burnt in 1000 cubic feet air. Maximum theoretical concentration is 5%, but 4% more typical.

For building fumigation, some authorities recommended sulfur burnt to yield 4% concentration for 10–12 hours. (Bishopp 27; Creel 1901) There would probably be a separate sulfur pot for each room of the building. For ship fumigation, the standard exposure was with 3% gas for five hours. (Treasury 212).

Sulfur may be burned in a furnace (perhaps 1800°F), with induced draft. Sulfur trioxide, visible as white fumes, is also produced. The sulfur oxides are drawn through water-cooled tubes into a blower. By displacement of air it is possible to achieve higher concentrations of sulfur dioxide in the fumigated compartments. The Clayton furnace was once popular for fumigation of large vessels.

Sulfur dioxide may be liquefied by compression (to four atmospheres) or cooling (to -18°C) and then transported to the fumigation site in pressurized or refrigerated tanks. Liquid sulfur dioxide can be vaporized without risk of fire. Two pounds of sulfur dioxide correspond to one pound dry sulfur, and in 1917 the liquid cost ten times as much as sulfur.

Sulfur dioxide is quite toxic to insects but unfortunately “has a deleterious effect on grain and flour and is highly corrosive to metals.” (FAO). There are authorities who complain of its lack of “disinfectant” (germicidal) action (Simpson 390) by the older sulfur fumigation methods, but contrary to early-twentieth century belief, contact with surfaces bearing the free bacterium don’t play an important role in transmission to humans.

Carbon monoxide is generated by the incomplete combustion of carbon. On steamships, funnel gas (a mixture of carbon monoxide and dioxide) has been used for fumigation against rats. However, it’s not toxic to fleas (Wardle 156) unless combined with dry heat (Pollitzer 497).

Cresol (carbolic acid, phenols) can be isolated from coal tar. It’s gently heated to vaporize it, and is toxic to insects at a concentration of 4 ounces per 1000 cubic feet. I don’t know whether it has any effect on rats; it’s probably more suitable as a room fumigant for lay use. (Wardle 157).

Nicotine (F, HTA, boils 247oC) was first used as a greenhouse fumigant in 1825, by burning tobacco. In France it was used as a tent fumigant; wood shavings soaked in tobacco juice were suspended in a wire basket and set on fire. (Wardle 159). In at least 1630s England and the Netherlands, grave diggers, corpse carriers and physicians, such as Doctor Ysbrand van Dimerbroek of Nijmegen, deemed the tobacco pipe to be useful as a portable fumigation device, protecting them from the noxious vapors thought to carry the plague. (Byrne 121, 344) While that would not be useful against the real threat—rats and fleas—from there it is not too great a step to burning the tobacco on a brazier so as to purify an entire room. Of course, you would still need to recognize that the necessary concentration and exposure were high enough to endanger humans, and take appropriate precautions.

Naphthalene is a coal tar constituent, and was once the active ingredient of mothballs. It’s a solid that sublimes at room temperature. There has been debate as to whether it’s truly toxic or merely an insect repellent. (Timar-Balazsy 293).

Hydrogen cyanide (F, LTA, boils 26°C) was first isolated in OTL 1786, although bitter almond extracts were used as poisons well before RoF. It was used to protect orchards against scale insects in 1886. In 1915 it was the cheapest of the fumigants, thanks to its high potency.

It can be generated on site by reaction of sodium cyanide with sulfuric or hydrochloric acid. One fluid ounce of cyanide and one of acid, to three of water, is sufficient to fumigate 100 cubic feet. The fumigation should be for at least 4–6 hours, preferably a full day. (Howard).

To permit it to be transported and stored safely, in the 1920s it was absorbed into either diatomaceous earth or gypsum pellets. In this form, it was called Zyklon B, and in WW II became infamous (with the warning odorant omitted) as the gas used to exterminate the Jews.

Calcium cyanide reacts with moisture in the air to generate hydrogen cyanide. WHO says to use it only outdoors.

I’d be very reluctant to recommend use of hydrogen cyanide in any form until the antidote (amyl nitrate) is also available. Even then, I think I’d urge that its use be limited to ships that can’t be unloaded, fumigation chambers for cargo, and isolated food storage facilities.

Hydrogen phosphide (F, HTA) has been used to fumigate insects in stored products, and rat burrows.

Methyl bromide (NF, HTA, boils 3°C) came into common use as a fumigant in the 1930s. It is nonflammable and nonexplosive under ordinary circumstances, and its vapors dissipate rapidly. It is very good at penetrating into porous materials at ordinary pressure, but it is not as toxic as hydrogen cyanide. (Bond).

Carbon disulfide (F, HTA, boils 46°C) was used as a soil fumigant in 1869. It’s a liquid that evaporates quickly when exposed to the air (boiling point 115°F). It can be ignited by an open flame and it has been implicated in many explosions (FAO). Hence, it’s considered too dangerous to use in buildings. (Pollitzer 496). Early-twentieth century Public Health Surgeon Rucker considered it “one of the best of all weapons” against rural rats and squirrels. You’d saturate a pad of absorbent cloth with the liquid and thrust it into a rodent burrow (after, presumably, blocking all known escape holes). (Treasury 158).

Carbon tetrachloride (NF, HTA, boils 77°C) is the poison that every amateur entomologist is familiar with. However, as shown later, it is of relatively low toxicity.

Chloropicrin (NF, boils 112°C) was first prepared in 1848, and was a byproduct of the explosives industry (it’s a picric acid derivative). It was used in WW I for chemical warfare. It’s not as potent as hydrogen cyanide but it has been added to the latter as a warning gas.

Sulfuryl fluoride (NF, HTA, boils -55°C) was first used in the 1950s to kill termites. It came into common use as a structural fumigant. It’s odorless so a warning gas should be added.

FAO recommends use of hydrogen cyanide, methyl bromide, and chloropicrin for rat control in empty storages and warehouses, hydrogen cyanide or methyl bromide in ship holds; just methyl bromide in spaces containing food, and hydrogen cyanide or phosphine in outdoor burrows and garbage dumps. (Bond Schedule T).

The Grantville up-timers will not know the relative toxicity of the various fumigants to rats and fleas, but this can be determined by experiment. According to available rat LC50 data (LC50 is the concentration that kills 50% if exposed for the stated period, typically 1 or 4 hours), it appears that in terms of effect against rats, hydrogen cyanide > methyl bromide > sulfur dioxide > sulfuryl fluoride > carbon disulfide. However, when the rats have the opportunity to flee, other factors than simple toxicity come into play.

In 1915–16 New Orleans, public health officers determined the efficiency of fumigation with either sulfur dioxide (3 pounds per 1000 cubic feet, 6 hour exposure) or hydrogen cyanide (5 ounces per 1000 cubic feet, 0.5 hours for superstructures and 1.25 for holds). The efficiency was the number of rats killed by fumigation as a percentage of the sum of those so killed and those trapped after fumigation. The results were as follows:


Target Sulfur Dioxide


Hydrogen Cyanide


entire vessel, loaded or empty 77 95
loaded holds 64 80
empty holds 96 99
above-deck compartments 55 94



This showed that sulfur dioxide was excellent for fumigating empty holds, but otherwise inferior to cyanide.

Insofar as efficacy against fleas is concerned, data are very sketchy, but it appears that hydrogen cyanide > methyl bromide > chloropicrin, carbon disulfide > sulfur dioxide. (Pollitzer 526).


Ships, warehouses, food storage facilities, food containers and residential buildings can all be fumigated.

To be effective, the target area must be sealed so the fumigant cannot readily escape. In addition, if it is possible to deprive the rats in advance of possible refuges, that’s advisable.

Creel suggests that the fumigation may be validated “by placing captive fleas in a test tube within the room and rats in cages well protected by loose covering to simulate natural harborage. A small amount of dust in bottom of the test tube will simulate natural conditions for fleas.” If the rats and fleas die, that’s promising . . . .

Ship Fumigation . Absent detailed ship plans or deliberate measurements, you can assume 1000 cubic feet for every ten net tons. I would expect wooden hulls to be leakier than metal ones, and Havard (767) says that with 5% sulfur dioxide, you want 48–72 hour exposure in wooden ships as opposed to just 24 hours in metal ones. Creel (1900) thinks that 4% sulfur dioxide for 12 hours suffices if the vessel is empty.

The ship must be prepared for fumigation by removing dunnage, ropes, canvas, etc. and opening up dead spaces such as bilges, double sheathing and piping. It’s nice if the ship is designed so these spaces have removable or hinged access panels. (Grubb 1268).

Building Fumigation . If the walls themselves were airtight, then it might suffice to close and caulk windows, and from the outside place gummed paper strips over the doors and windows. (Howard). It may also be helpful to paper over the floor edges. (Hinds). However, I suspect that the walls, floors and ceilings are leaky, and that it will be necessary to “tent” the entire building with a flexible, gastight material.

This will not be easy in the 1630s; even with the changes wrought by the RoF, rubber will be in short supply, and plastic sheets even more so. At this point, it appears that the most effective strategy would be to cover the building with sail cloth. While sail cloth is intended to “catch the wind,” and thus has some resistance to gas transfer, it isn’t totally impermeable, either. Indeed, a test showed that “10 ounce duck, untreated, will transmit practically 85 per cent as much air under a slight pressure as though there were no such obstacle in the way.” (Hinds 3). So several layers may be needed to reduce the permeability to an acceptable level. It’s also possible that it would be advantageous to use some kind of varnish, as was done with certain nineteenth-century balloons.

A further complication is that few of the individuals in the part of town at greatest risk (i.e., highest rat concentration) are likely to live in single-family buildings, and the deadly fumigant might percolate out through the party walls. It’s possible that effective fumigation will require working on a block-by-block basis.

If the building is multi-level, then with a lighter-than-air fumigant like hydrogen cyanide, you start at the top and work your way down, and with a heavier-than-air one like carbon disulfide, you do the reverse. (Hinds 7).

Next the fumigant must be introduced into the target area. Either the fumigant in active but enclosed form must be transported there, and then released, or it must be produced in situ by chemical reaction. If the release or in situ production is triggered manually, the operator must wear suitable protective gear (respirator). The design of respirators is beyond the scope of the article, but for now I can say that these will probably either involve absorption of the gas onto activated charcoal, or inactivation of the fumigant by reaction with a chemical agent. It will be necessary for the respirator to include a reservoir of fresh air if the contaminated air can’t be purified.

Not only is this problematic from an infrastructure standpoint, when the operator exits the target area, there will necessarily be some release of the fumigant. Hence, I think that the release or production would have to be on some kind of time delay, e.g., a fuse or clockwork mechanism, so that the operator can get clear and complete the seal before it occurs.

Temperature affects efficacy; at low temperature, the fumigant vaporizes and diffuses more slowly, and insect metabolism (and thus respiration vulnerable to attack) is retarded. At high temperatures, the fumigant may dissipate too quickly. It’s also advisable not to fumigate on a windy day.

The building must remain sealed for sufficient time for the fumigant to do its work; this period will need to be determined by trial and error. Effectiveness may be judged by exposing test insects and rodents to the fumigant in a suitable confinement.

Finally, the building must be cleared of the fumigant without endangering the health of people in adjacent areas. It’s possible that the miner’s canary will come in handy for testing the air quality.

Ultimately, I have my doubts as to how well fumigation will work for period residential buildings. Even in Rucker’s day, in buildings “it was difficult to confine the gas by stopping all cracks and other openings.” Rucker thought that the use of fumigants in the city was “limited chiefly to warehouses, elevators, and ship holds.”

I am also very concerned about the hazard to human health posed by fumigants. Even in a tech-savvy culture, they are best left to well-trained professionals.

It has been found that people will fail to report finding rats in their homes because of the stigma of having the home fumigated. (Treasury 233).

Supportive Care

Generally speaking, the patients should be transported to the hospital if not cared for at home (those who walk to get treatment tend to have poor outcomes) and stay in bed until the disease is brought under control with pharmaceuticals, If they get up and about, even to relieve themselves, there’s a risk of heart failure. Since fever is one of the symptoms, fluid intake must be maintained. A light diet seems best. It’s definitely a bad idea to excise the buboes. (Pollitzer8, 117).

Prophylaxis and Treatment in Humans: Introduction

Topical Treatment. In the seventeenth century, it was not uncommon for buboes to be lanced and drained. This was a bad idea because it would expose the caregiver to infectious material. However, a bubo may be aspirated to take a sample for diagnostic purposes, or to relieve pain and swelling.

In the late-nineteenth and early-twentieth centuries, there was some experimentation with application of topical antiseptics (iodine, mercuric chloride, carbolic acid, potassium permanganate) to or even their injection into the buboes, without much success. (Jennings 127).

Systemic Prophylaxis and Treatment. In the late-twentieth century, the accepted plague therapy was with antibiotics or other chemicals. Antibiotics were also used for post-exposure prophylaxis. Less established therapies include vaccines, antisera, phage and bacteriocins.


Antibiotics are the chemical weapons employed by fungi (especially Actinomycete) in their ceaseless war with bacteria. Scientists identified particular fungi as producing antibiotics because of their visible suppression of bacterial growth in culture. They then isolated the active agents and characterized their structures. If we were lucky, the structures were simple enough that the compound could be synthesized chemically. And chemists could also suggest modified structures and test them for improved activity (greater potency, less specificity, greater safety).

If not, the compound could be obtained only by growing the mold and isolating its antibiotic product. The compound could be improved upon only by mutating the producing organism and hoping for a change in the antibiotic, or by chemical modification of the natural antibiotic (semi-synthesis).

In Grantville literature, the modern EB says, “Treatment is primarily with streptomycin, tetracycline, and sulfonamides. Penicillin is without effect against plague.”

Streptomycin is the modern drug of choice, when available, for treatment of plague. It’s produced by the soil organism Streptomyces griseus , and the organism was first isolated OTL 1943, from “leaf compost, straw compost and stable manure” on the Rutgers College of Agriculture farm. (Pringle). The antibiotic’s structure was determined in 1948. Its NTL availability is dependent on either fortuitously isolating the same organism (very unlikely, given the likely scale of a 1632 universe screening program, and you’d also have to recognize that you had S. griseus in hand) or synthesizing it on the basis of a structure published in Grantville literature (also unlikely; it’s an aminoglycoside, a rather complex compound, and OTL its total synthesis didn’t come until 1974) (Umezawa).

Another modern first-line drug is gentamicin , another aminoglycoside (actually a mixture of three); same problems. The polyketide tetracycline, and the related drugs oxytetracycline and doxycycline, have been recommended for plague therapy in recent years (although FDA says that there have been no controlled studies comparing them to the aminoglycosides). None of these are likely to be available in the NTL 1630s, either from their original biological sources (hard to find) or by synthesis (too complex). See Appendix.

Of the first-line anti-plague drugs, the only one that will be available in the NTL 1630s ischloramphenicol . Canon says that chloramphenicol had already been produced by September 1633, and possible synthetic strategies are discussed in Cooper, “Industrial Alchemy: Part 3, Organic Chemistry Methods and Canonical Appearances” (Grantville Gazette 26).

In modern medical practice, chloramphenicol is given 500 mg intravenously every 6 hours, for fourteen days (as therapeutic; Webber 246 says 7–10 days) or 7 days orally (as post-exposure prophylactic). Its use is especially recommended when plague meningitis has developed. (Torres 475). It has been successfully used (in the 1940s) to treat pneumonic plague. (Pollitzer8, 114).

Canon has thus far not paid much attention to the difficulties of intravenous administration in the NTL 1630s, i.e., hypodermic needles are new and probably in short supply, and medical appreciation of antisepsis is also limited. It’s clear that the NTL doctors will try giving chloramphenicol orally, even for treatment, if intravenous administration isn’t possible, and it turns out that it was so used back in the 1940s. (Pollitzer8, 114). The other possible oral anti-plague agents are fluoroquinolone antibiotics (ciprofloxacin, levofloxacin, ofloxacin). (Brouillard). These are less daunting synthetic targets than the aminoglycosides or polyketides, but they are certainly more complex than chloramphenicol.

Sulfanamides . Grantville has unspecified “sulfa drugs” by Fall 1633. (Flint, 1633 , Chap. 35). These probably include sulfanilamide, whose synthesis is discussed in “Industrial Alchemy, part 3.” Grantville Literature also provides synthetic guidance for sulfacetamide, sulfadiazine, sulfadicramide, sulfaguanidine, sulfamidochrysoidine, sulfathiourea, sulfadiamine, sulfonylbisacetanilide, sulfoxone and the heterocyclics sulfamerazine, sulfamethazine, sulfamethizole, sulfamethoxazole, sulfamethoxypyridazine, sulfapyrazine, sulfapyridine, sulfaquinoxaline, sulfasomizole, sulfathiazole, and sulfisoxazole.

Prontosil was first used therapeutically against bubonic plague in 1938. (Pollitzer8, 104). It was discovered to actually be a prodrug, metabolized by the body into sulfanilamide, the first sulfa drug. Meyer constructed a “mouse village” into which he introduced infected fleas; the mice in the section receiving sulfanilamide lived and those that just had a normal diet died. (Sabin 305)

The sulfonamides are bacteriostatic—they stop reproduction rather than reducing the bacterial load. Clinically, large-scale use began in 1940–41 with sulfapyridine and sulthatiazole. Later, sulfadiazine and its derivatives became more popular. (Pollitzer8, 105). Sulfadiazine treatment saved 3 out of 5 pneumonic plague patients in a 1946 epidemic. (FDA 25).

In a mouse study, Sokhey demonstrated the efficacy of several sulfa drugs. He treated at either 48 hours after challenge (deemed prophylactic) or 72 hours (most likely after septicemia, which is at 48–72 hours in mice), with these results:


mouse % reduction in mortality 48h 72h
sulfapyridine 0% 0%
sulfathiazole 90% 70%
sulfadiazine 50% 40%
sulfamerazine, sulfamethazine 70% 70%


In 1941–48, in humans with bubonic plague, mortality rates were:


pre-bacteremia post-appearance of bacteremia
sulfapyridine 3.3 50.0
sulfathiazole 5.8 42.0
sulfadiazine 2.5 21.0
sulfamerazine 2.2 31.8
Iodine (control) 16.2 90.7


(Pollitzer8, 106).

By way of comparison, with streptomycin, the mortality rate was 1.8% pre-bacteremic and 10.8% post-bacteremic.

In 1952, WHO recommended an initial dose of 10g, and a total dose of at least 50g over the first week. (Id.). With sulfadiazine and sulfamerazine, the initial dose was intravenous and subsequent doses, oral; sulfamethazine (deemed equivalent in efficacy to sulfadiazine) was given only orally.

The sulfonamides have only a small niche in modern clinical practice, although the WHO Plague Manual still lists sulfonamides for use against plague in humans—for sulfadiazine, a loading dose of 2–4 g then 1g every 4–6 hr for 10 days. However, they are still widely used in veterinary practice.

There are two concerns I have with the sulfonamides, the first being that with these doses there might be significant side effects, and the second that there will be temptation to reduce the total dosage so more people can be treated.

On the first point, it has been suggested to me that “at least some of the adverse actions of the stronger sulfa medicines can be reduced by concomitant administration of folinic acid (which mammalian systems can use in folate metabolism, while microbes can not).” (Krin).


Plague vaccines present plague antigens to elicit activation of T-cells and production of anti-plague antibodies in the recipient.

For prophylaxis, they have the advantage that by administering just a few doses one can provide prolonged (how long, see below) future protection, whereas with a pharmaceutical you have to keep giving the drug on a daily basis. Like pharmaceuticals, vaccines have side effects, and thus the therapeutic benefit must be weighed against the adversities. Moreover, “protection is delayed for at least 1 week after immunization . . . .” (Anisimov 1462).

Vaccines are useful only prophylactically; they don’t cure one who has already contracted the disease. Indeed, there are reports of plague vaccines actually worsening the disease, possibly because they boosted the exposure to the bacterial endotoxins. (Chase 49). However, there is no short-term increase in susceptibility to plague. (Pollitzer 532).


Vaccines are not completely unknown to down-timers. A crude form of vaccination against smallpox was used pre-RoF in Asia and occasionally (mostly as a folk remedy) in Europe; this was variolation, which used a dried and thus hopefully attenuated (weakened) form. It is this method that is initially adopted by the Grantville Sanitary Commission, once its border screening brings someone with an active case of smallpox into their hands. (Offord and Coljee, “A Gift of Blankets,” Grantville Gazette 10.)

Several vaccines have since been developed with up-time guidance. In NTL 1633, the Grantville Sanitary Commission sends a mission to England to find a “cowpox” to use to cross-immunize against smallpox. It goes upriver from Bristol to Gloucester, and finds a horse with “grease” (horsepox). On the way back to Bristol it distributes crude horsepox vaccine with instructions on how to use it. (Offord, “Cowspiracy,” Grantville Gazette 12.) And catpox is used similarly in the Spanish Netherlands. (Flint and DeMarce, 1634: The Bavarian Crisis , Chapter 69.)

The tetanus death of an up-time child precipitates the formation of a company (Daisy Matheny Biolab) for vaccine development. Its first vaccine (mid-1634) is a typhoid vaccine. The company is hopeful that it will enter human trials of the tetanus vaccine in 1635. (Offord, “Little Angel,” Grantville Gazette 10).


So, what about a plague vaccine? It won’t be known in Grantville, but Y. pestis has little genetic variability, and thus is an easy target for immunization (unlike, say, influenza) (Carniel 144). Historically, the problems have been reactogenicity (side effects), transient protection, and failure to protect against the pneumonic form.

The modern (as of RoF) EB mentions the availability of a plague vaccine; production was discontinued in 1999.

The first step to making a vaccine is isolating the pathogen, and that should be easy enough; there are millions of germs in each bubo. The catch is handling the pathogen so as not to contract the disease. Nowadays, a biosafety level III containment facility is required because of the risk of infection by inhalation. That means a double set of self-closing locked doors, steam sterilization of discarded cultures, protective coats, gloves and glasses, handwashing, negative pressure ventilation, medical surveillance of personnel, easily decontaminated furnishings, and a biological safety cabinet with inward and outward HEPA filters.

Of course, the early plague researchers did not have such facilities, so it is possible to develop a vaccine without them. It is also possible to get sick as a result.

Even if a vaccine is developed in the NTL, there is no guarantee that it will be effective for a long enough period to be useful, or that its protection will outweigh its side effects. OTL history demonstrates just how difficult it is to make a truly useful plague vaccine.

Immunization with pus from a bubo was at least proposed by Weszpremi (Hungary, 1755) and Samoilovski (Samoilowitz?; Russia 1781). Whyte (England 1802) inoculated himself and four assistants and all died a week later. There were other instances in which the inoculation had no effect at all, probably because the inoculum was from a convalescent and non-infectious. (Pollitzer3, 185; Huygelen 93).

The Haffkine vaccine was widely used in 1897–1919, and is referred to in EB11/Plague without particulars other than that there was a controversy as to the inclusion of carbolic acid. It was a preparation of heat-inactivated bacteria (KWC: killed whole cell), injected subcutaneously.

I believe that the bacterium initially was cultured for 5–6 weeks and then killed by exposure to 70oC for one hour. (Bazin 344). However, the treatment later became 55oC for 15 minutes. (Pollitzer3, 190). During the SF epidemic, Haffkine was able to supply 10,000 doses a week. (Chase 49).

In 1900 Honolulu, the vaccine was found to vary in potency, which meant in turn that doctors had to guess at the proper dosage. (Mohr 165). It took a week to develop immunity, and protection lasted 6–7 months. (Simpson 332).

In India, according to Haffkine, the mortality in the vaccinated population was 1.6%, and in the non-vaccinated, 24.6%. (Bazin 344). However, we don’t know whether the two groups had the same exposure to the disease.

The usual side effects were fever, headache, nausea, malaise, pain, and local swelling; these were incapacitating for a day or two and persistent for a week or more. (Simpson 331). When the Haffkine vaccine first arrived in Honolulu, two doctors publicly injected each other. “Both men became seriously ill” for several days. (Mohr 177–8). A SF reporter experienced the following side effects: shooting pains and numbness in the inoculated arm, dizziness, tinnitus, fever and stupor, overall lasting eight hours. (Chase 49). There were mass public protests against a vaccination order. (59). The severity of the vaccination reactions was so great that in one of the Pacific Islands the authorities discontinued its use. (Meyer).

There were experiments with other inactivating agents (formaldehyde, merthiolate, sulfathiazole, alcohol), culture media (casein hydrolysate), incubation temperatures, etc. (Pollitzer3, 190ff).

The modern plague vaccine comprises bacteria grown on agar and killed with formaldehyde. The immunization protocol required a booster 1–3 months after the initial dose. Surprisingly, no clinical trial has demonstrated its efficacy even against bubonic plague. However, there’s circumstantial evidence; a version was used in Vietnam, and the incidence of bubonic plague in servicemen was 8 case per million, versus 333 in unvaccinated Vietnamese. (But bear in mind that the US Army had much better sanitation than the Vietnamese did, and so probably less exposure to infected fleas.) Significant side effects were common and could be incapacitating. (Meyer: Barrett 1089; Titball).

According to the 1988 directions for use, it was then given intramuscularly, as two or three injections. The second injection is 1–3 months after the first, and the third 3–6 months after the second., “The duration of protection against infection following administration of the primary series of these injections of Plague vaccine is brief (i.e., 6–12 months) and booster doses in approximate 6 month intervals are required for continued protection.”

Whatever its effectiveness against bubonic plague, the KWC vaccine didn’t help much versus pneumonic plague. For example, a chemist at Walter Reed, who had received vaccination in March and on Aug. 10, 1959, was exposed on Aug. 29 and displayed symptoms on Sept. 1. (FDA 26). And mouse studies have shown a failure to protect against inhaled Y. pestis. (Titball).


The side effects of the KWC vaccine can be reduced by use of selected fraction (subunit) of the organism or even better a purified antigen. That requires a higher development of biotechnology than just isolating and killing the bacterium.

I was surprised to discover that a “nucleo-protid” vaccine was reported by Lustig and Galeotti in 1897. The bacteria were killed with potassium hydroxide and then the vaccine was made by filtering off the cell debris, precipitating the supernatant with hydrochloric or acetic acid, and dissolving the precipitate in a weak solution of sodium carbonate. This no doubt impure product was toxic but in a dose of 3 mg in humans just caused fever, malaise and local swelling. (Simpson 409; Lustig; Pollitzer3, 172). It was used therapeutically in 1898 Bombay and considering all cases, the treatment increased the chance of recovery by 55%, and if both “moribunds” and “semi-convalescents” were excluded, by 96% (to about 40% of the serum-treated cases). (Choksy).

According to Bergey’s Manual of Determinative Bacteriology (8th edition), Baker (1947) was the first to isolate the F1 fraction. This was probably a rather impure extract, as another source says that the first Yersinia protein to be isolated was the secreted 17-kDa capsular F1 antigen, in 1974. (Carniel 72).

Until the 1632 universe advances to the stage of being able to sequence proteins or genes, and to make and express recombinant DNA, F1 can only be obtained by culturing Y. pestis, which obviously presents workplace health risks. (Skurnik 399).


An attenuated strain of Y. pestis (EV 76 Madagascar) was developed in 1908 and brought into large-scale use in 1934. (Pollitzer3, 197). It has been reported to provide greater protection against both bubonic and pneumonic plague than the formaldehyde-killed vaccine in mice. (Russell). It is not avirulent; in mouse studies, the fatality rate has been 1%. (Titball).

The attenuated strains normally had to be used within one month of their preparation. There is controversy as to whether they can be preserved longer by freeze-drying. (Pollitzer3, 201).

Attenuation of virulent bacteria is more art than science, and the medical personnel in Grantville are unlikely to have more than a modicum of theoretical knowledge about the process. In essence, it involves passaging the bacteria through different culture media or animal hosts, then testing for virulence. In practice, plague strains were attenuated by cultivating virulent strains under hostile conditions (in immune animals, under high or low temperature, in the presence of alcohol or anti-pestis phage, etc.). (Pollitzer3, 170).


Cross-immunization . There are two other human pathogenic Yersinia species, Y. pseudotuberculosis and Y. enterocolitica . All three species share the V antigen, which in 1963 was shown to be a protective immunogen. It has been demonstrated that mice are partially (50%) protected from bubonic (not pneumonic) plague by prior exposure to Y. enterocolitica (Alonso 1980) and pseudotuberculosis. (Wake 1978; Pollitzer3, 193). There is both a short-term (antibody) and long term (T cell) protective immune response. It turns out that the V antigen is polymorphic (exists in multiple forms), so whether cross-immunization works depends on the similarity of the V antigens in question. (Roggenkamp).

Both pseudotuberculosis and enterocolitica can cause gastrointestinal upset and even death (although they are certainly much less malevolent than pestis). Some naturally avirulent strains (enterocolitica 4052, pseudotuberculosis IP32680) have been isolated, (Carniel 148), and so there has been post-2000 work in genetically engineering them to delete virulence factors. (Carmiel 154).

There are at least ten other Yersinia species that don’t cause human disease, but they were only identified in the 1980s or later and there have been no cross-immunization studies employing them.

Another interesting tidbit is that Yersinia produce bacteriocins, which suppress other bacteria, even other Yersinia species. On the one hand, this means that cross-immunization isn’t likely to be effective in someone already infected with Y. pestis , as the pestis bacteriocins will inhibit the competitor. On the other hand, those bacteriocins might themselves be of therapeutic value. In 2012, it was found that a Y. frederiksenii colicin Fy inhibited pathogenic Yersinia . (Bosak).


A vaccine comprises immunogens that stimulates the host’s immune system to produce antibodies and to activate T-cells, thus providing defense (acquired immunity) against the infection. Alternatively, one may immunize an animal host, and extract the antibodies from the animal’s blood serum; these extracts are called antisera. This may be advantageous if an animal responds more vigorously to the available immunogen than does a human.

Since the antisera comprise foreign (animal) proteins (the antibodies), there is a risk that if they are administered to a human, the recipient will experience an immune reaction. This risk can be reduced by use of an antibody fragment rather than whole antibody. Papain digestion (yields Fab) is from papaya, which is not especially available in Europe. Pepsin yields (F(ab’)2 and is much more available. Reduction yields Fab’.

Typically, an immunized animal will continue to produce antisera for several months following the last injection. Eventually production will peter out and a new immunization is needed. If the immunogen is the killed bacterium, then that means handling a dangerous organism again.

Antisera may also be obtained from human volunteers who were immunized repeatedly to build up their antibody levels, or who survived the infection. Of course these sources tended to be in limited supply. Nonetheless, in NTL 1632 Dr. Sims is treated for smallpox with such anti-serum. (Offord and Coljee, “A Gift of Blankets,” Grantville Gazette 11).

There were a couple of antiplague antisera (i.e., polyclonal antibodies) used before antibiotics. The first was the Yersin rabbit antiserum (1895). Yersin also developed a horse antiserum (1896). There was Naidu’s buffalo antiserum (1931) and Sokhey’s horse antiserum (1936). For production, the horse has the advantage of size (more blood) and the rabbit of being susceptible to the bacterium (higher response). (Pollitzer3, 204).

Yersin’s horse antiserum was routinely used for both prophylaxis (in physicians only) and treatment during the SF epidemic. (Chase 49). Since it contained nonhuman proteins, there was the danger of “serum sickness” (an allergic reaction). This can be life threatening, but in general the horse antiserum was safer than the Haffkine vaccine. (Chase 83). The usual side effects were rheumatism-like symptoms. Protection only last two weeks, but patients were more willing to received repeat dosings. (Simpson 332).

It’s somewhat confusing, but it appears that Haffkine also developed a therapeutic “serum” (antiserum). In 1941–48, the mortality rate was 23.5% overall in bubonic patients treated with this serum, 50.7% if they had already exhibited bacteremia (bacteria in the blood), and 1.2% if they hadn’t. (Pollitzer8, 103). These results are markedly inferior to those achieved with the sulfanilamides (let alone streptomycin), and that helps explain why antiserum therapy was eclipsed by drug therapy.

Other Biologics

Phage. Phage are viruses that infect bacteria. Since phage reproduce in the bacterial host, a relatively small dose can be administered, which reduces the chance of side effects.

Phage are also very specific in their action. This is advantageous in that it further reduces the chance of side effects. It is disadvantageous in that you aren’t going to find a broad-spectrum anti-bacterial phage, you need to keep looking until you find an anti-plague phage and the possibility exists that the phage will not lyse all strains of Y. pestis. (Rashid, Table 1).

Bacteria develop resistance to phage as well as to antibiotics, but Carlton says (without attribution) that bacteria mutate to resist antibiotics once in 10^6 divisions and vs. phage once in 10^7 divisions. However, that may be because in the modern world, there is much more of a pool of antibiotic resistance genes, and so it might not be true early in NTL. The problem of resistance may eventually be addressed by mixing phage that bind to different bacterial receptors, or finding phage that target a pathogenicity factor (so a mutation that protects against the phage may also expose the bacterium to the immune system), but in early NTL we won’t be able to screen phage for their specific biomolecular targets.

Phage therapy has certain problems, some of which will be evident to Grantville physicians:

1) you need to work with the active pathogen, not only in identifying phage that are effective against them (that’s true of antibiotic screening too) but also in producing the phage—you are growing the phage in those hosts;

2) thus, there is a need to ensure that the phage preparation does not include live pathogen (without denaturing the phage);

3) the high specificity of phage means that either the pathogen in a patient must be typed (perhaps not just to species, but also to strain), which would be difficult and time-consuming in 163x, or you must administer a phage cocktail;

4) phage can be cleared rapidly from the body;

5) patients can develop anti-phage antibodies—phage are more likely to be immunogenic than antibiotics, especially small molecule antibiotics such as chloramphenicol; and

6) lysis releases bacterial endotoxins. (Pollitzer3, 216).

In 1925, d’Herelle successfully treated four plague patients with a highly virulent anti-plague phage isolated from rat feces. However, subsequent studies yielded conflicting results. (Anisimov).

It appears that at least in some cases, researchers used harsh treatment (heat, oxidants, etc.) to inactivate any surviving bacteria and that these treatments also denatured the phage coat proteins. Hence, a negative result might in fact have been attributable to the lack of functional phage. Also, some old phage trials used lysogenic phage—phage that can replicate without killing the host cell. What we want are lytic phage, those that rip open the cell to release their progeny. On the other hand, some of the positive results were in studies that did not feature proper controls.

There have also been experiments with mixtures of phage and plague bacilli. (Pollitzer3, 217). However, I’d be worried about the release of bacterial endotoxins.

I doubt that the physicians in Grantville are going to know much about phage therapy, for any bacterial disease let alone plague, that having been popular mainly in the Soviet Union. They should however know that phage infect bacteria and expect to find phage in bacteria-rich environments.

I was pleasantly surprised to discover that anti-pestis phage could be found, not only in the blood and tissues of plague-infected humans and animals, but also in soil and water. “In 1927 Flu (22) recovered phages from canal waters in Leyden, Holland, that were active against Y. pestis, Escherichia coli, and Shigella dysenteriae.” (Garcia). Phage “PY100 was isolated from pig manure collected on a farm in Germany.” (Schwudke). In Athens, at a time when plague didn’t appear to be active, 2/217 rats had blood containing anti-pestis phage. (Pollitzer3, 215).

Bacteriocins. Bacteriocins are the bacterial equivalent of antibiotics, chemicals they produce to kill competing bacteria. The problems of developing them are also analogous. Also, they tend to be narrow spectrum and cytotoxic.

Therapeutic Considerations

Yersinia pestis strains with multidrug resistance were isolated in Madagascar in the 1990s. (Abbott 56). Of course, that was after streptomycin had been in use for half a century.

My fear is that because the drugs will be expensive and in short supply in the 1630s, there will be a temptation to reduce the total dose per patient. That may lead to relapse, and it will tend to breed a resistant strain.

Some medical sources contain dire warnings to the effect that antibiotic therapy must be commenced within 24 hours of appearance of symptoms or death will occur, especially in cases of pneumonic plague. If these warnings appear in Grantville Literature, and are taken literally, a physician facing a supply shortage might well decide not to “waste” the antibiotic on someone belatedly diagnosed with plague.

Delay in treatment is of course a concern. A particular problem is that some of the drugs are bacteriolytic, that is, they rip open the bacterium. That releases bacterial endotoxins, mimicking septicemic plague and perhaps causing septic shock. The greater the delay, the greater the bacterial population, and the more endotoxin release.

Clearly, there’s a falloff in effectiveness as treatment is delayed. A study of mice challenged with pneumonic plague showed that streptomycin was 100% effective when given after 24 hours, and 59% after 42 hours. (Byrne). And an account of the fight against pneumonic plague in Surat 1994 said that those treated within six hours usually recovered, those reached after twelve hours had a fifty-fifty chance, and those treated after 24 hours were unlikely to be saved. (Marriott 52).

However, that begs the question of when the clock started ticking. When the first symptom appeared? When plague was diagnosed? And it’s likely that there’s more time to act in cases of bubonic plague.

Control of Transmission from Wild Hosts to Urban Rats or Humans

While wild animals will generally try to avoid people, environmental change can cause them to change territory. Plague outbreaks have been associated with flooding in India (1994) and Zambia (1996), and drought in Mozambique (1994). (Abbott 16).

The destruction or relocation of a wild rodent plague focus is likely to be extremely difficult, and the attempt can backfire if the rodents flee toward civilization. (Webber 246).

In general, people should be warned to stay away from large colonies of rodents, like prairie dog “towns” in the American southwest, and to be wary of lethargic or dead animals. Also, you shouldn’t “skin a dead animal without wearing gloves and a long-sleeved shirt.” (Fagerlund 98). If you must pass through a plague focus, then wear long trousers tucked into socks,” both treated with flea repellents or insecticides, and don’t eat or even touch the rodents.” (Webber).

Once more immediate health concerns are addressed it may be advantageous to monitor rodent colonies that are near towns or major road arteries to determine whether they bear infected fleas. Certain animals (notably coyotes, dogs, badgers, raccoons, skunks, and black bears) can be infected by plague, and produce antibodies against it, but are not sickened by it; they can be used as “sentinel” animals to detect that the plague is emerging from a natural reservoir. (Acha 215; Abbott 47).

In recent years, efforts have been made to introduce insecticides into wild rodent burrows in order to reduce flea populations (Id.), but this has been as much to protect endangered mammals as to safeguard human health, and it is labor intensive and time consuming. There has even been work on vaccines for prairie dogs and ferrets. (54).


I would predict that in some European communities, such as Magdeburg, Venice, Essen, and post-Baltic War Amsterdam and Copenhagen, there will be access to up-time rat and flea control agents, building materials, and human prophylactics and therapeutics, not many months after those are duplicated in Grantville itself. In others, whether because of remoteness from trade routes, economic or technological backwardness, or active ideological resistance to up-time influence, there will at most be knowledge (and not necessarily acceptance) of the up-time teachings concerning the role of bacteria, fleas and rats in the propagation of plague. And of course still others will fall in-between those extremes.

At the high end, I would expect the following: quarantine and fumigation of ships; rat-proofing of at least new construction; rat-catching for monitoring purposes; extensive use of DDT in rat-infested areas for flea control; stringent plague reporting requirements; bacteriological and agglutination testing of blood of human suspects (and perhaps also rats found dead); DDT treatment of persons and bedding of plague victims and “contacts”; isolation of pneumonic plague cases; treatment of at least bubonic plague cases with chloramphenicol and/or sulfa drugs; “prophylactic” drug treatment of pneumonic plague contacts (supplies permitting); and possibly vaccination of physicians and nurses with KWC vaccine. Ideally ship fumigation would be carried out by transferring the cargo to a custom-built, rat-proof fumigation facility equipped for hydrogen cyanide fumigation, and fumigating the empty ship with sulfur dioxide.

At the low end, limited to pre-RoF commodities, there are still measures to be taken. Quarantine periods can be reduced (which will be welcomed). Smaller buildings can be rat-proofed by elevation; new large construction built using brick or stone. There will probably be more extensive rat extermination by poison, traps and natural predators, with protective gear (clothing, tongs) used by those disposing of the corpses. Fleas will be trapped and there will be experimentation (alas, probably not controlled experimentation) to find substances that repel or kill fleas; some sort of hydrocarbon emulsion is perhaps the likeliest. There may be some attempt at sulfur dioxide fumigation of ships and houses. Those who contract the plague will most likely only receive supportive care. However, some experimentation with phage or antisera is possible. For prophylaxis, if chemical synthesis isn’t an option, the only recourse is to attempt to develop a killed whole cell vaccine, with the bacteria killed with heat or perhaps strong acid.

In fourteenth-century Italy, Gabriele de Mussis described the plague as “sharp arrows of sudden death.” (Byrne 23). Up-time science offers several shields against this scourge, but it remains to be seen how soon the governments of the new time line will be able and willing to take advantage of them.