Mania a Potu at the Philadelphia Alms House, 1828-1850

Note: this is a historical methods paper on studying the dt's (delirium tremens or mania a potu) at the Alms House in Philadelphia. I submitted it to Pennsylvania History as a companion piece to a much longer article I had previously submitted. It should also be noted that my application to be a community columnist at the local paper was not successful.

Here you go:

Mania a Potu at the Philadelphia Alms House, 1828-1850
Ric N. Caric
Morehead State University


1. Introduction.
Mania a potu was the most common ante-bellum name for a hallucinatory disorder that affects extremely heavy drinkers who suddenly abstain from consuming alcohol. Now known primarily as delirium tremens, mania a potu was most prominently characterized by frightening hallucinations of being attacked by murderers, having devils in the room, seeing the walls falling down and the like. Recognized as a public health problem as early as 1825, mania a potu was the subject of considerable comment by temperance activists and literary writers during the 1830’s and 1840’s. Melville warns against the dangers of mania a potu in White Jacket for example. Although there are few, if any, historical studies of mania a potu, there are extensive materials available for studying the illness in Philadelphia, including the records of the Philadelphia Alms House, Pennsylvania Hospital, the Register of Death, and several medical dissertations. The Alms House records could be an especially useful tool for investigating mania a potu among poor men and women. A much larger medical facility than Pennsylvania Hospital, the Alms House kept extensive records on between one and three thousand men annually beginning in 1828. Because the Alms House records included medical and demographic information on almost every patient, they could serve for studies on the individual medical histories of patients, success rates of mania a potu treatments, comparisons between mania a potu and other kinds of disorders, and the distribution of mania a potu by race, ethnicity, gender, age, trade, and marital status.[1]
However, the Alms House records will be of very little use until there is more clarity about what counted as a case of mania a potu. The difficulty is that there is a sharp discrepancy between the number of cases found by medical authors and those listed in the Alms House patient registers. Examining mania a potu at the Alms House for short periods, authors like John Prosser Tabb found many times the number of cases and deaths from mania a potu that were listed in the registers for Alms House patients in those years. Because the patient registers had listings for all those admitted to the Alms House, the findings of the medical authors imply that hundreds of men and women who suffered from mania a potu were listed as being admitted to the Alms House even though they were not listed as mania a potu cases. As a result, the information on mania a potu patients in the Men’s and Women’s Registers will not be useful for research unless a method can be developed for figuring out which patients were afflicted by mania a potu even though they were not being listed as a mania a potu case.[2]
This article will discuss the nature of the discrepancies between medical writing and the Alms House registers and suggest methods for bringing the data from the Men’s and Women’s Registers more into line with the medical articles. First, I will discuss the extent of the discrepancies and put forward several reasons why the Men’s and Women’s Register would have undercounted mania a potu cases. Second, I will suggest three strategies for reinterpreting Alms House alcohol and insanity cases as mania a potu even though they were not originally listed as such. Finally, I will compare three possible strategies for bringing the Alms House registers in line with medical writing and discuss the strengths and weaknesses of the strategies. The focus will be on male mania a potu cases with a frequent references to the Women’s Register. Even though there ultimately will be a degree of indeterminacy in interpreting Alms House cases, the interpretive strategies suggested here should make the Register of Males more useful for studies of mania a potu.
2. Defining a case of mania a potu
The primary direct sources for data on mania a potu cases at the Philadelphia Alms House were the Register of Males and the Register of Females. Begun in 1828, these registers served as general registers for all men, women, and children who were admitted to the Alms House. The registers contain four types of information.[3]
A. Current Stay at Alms House date of admission, condition, ward assignment,
subsequent illnesses during stay, release from doctor,
date of departure

B. History at Alms House number of times admitted

C. Demographic Information age, race, place of birth, marital status,
number of children, employment

D. Temperance status notable intemperance

Those patients listed as having mania a potu or one of its synonyms (delirium tremens, the horrors, insanity from drink, etc.) were almost always placed in the cells before 1835. After 1835, the cells and the lunatic asylum were separated and most mania a potu patients were placed in the lunatic asylum. Mania a potu was one of three designations for alcohol disorders. Patients could also be listed under categories like “intoxicated,” “intemperate,” or “drunk.” There was considerable overlap between those listed as having mania a potu at the Alms House and those listed as “intoxicated.” Many of those repeatedly listed as having mania a potu also could be listed as “drunk” or “intoxicated” on other stays in the facility. For example, the physician and public speaker Alexander Draper was admitted to the Alms House five times in 1838 for mania a potu, but also was listed as “intoxicated” or “drunk” on three other visits that year. The same was the case with other repeat mania a potu patients like Ephraim Albertson, Joseph Calhoun, Werner Denard, and Thomas Swain as well. Like those listed as having mania a potu, those listed as intoxicated could be placed in the cells. They also could be placed in the medical ward or long ward.[4]
The Register of Males also had a category of “cause of poverty” under which Alms House personnel often listed patients as “intemperate.” Indeed, the most common comment made under this category concerned the temperance status of patients. The majority of those listed under the category of “cause of poverty” were listed as intemperate, with relatively few notices on patients being transferred to the Alms House from prison, “regular elopers,” and pay patients. There were also comments about the status of men whose appearance record-keepers found especially noteworthy. For example, Samuel Mulligan, a shoemaker, was described as “intemperate—ragged [and] broken down” upon admittance on March 30, 1841. Likewise, Samuel Porter was characterized as “intemperate” and “a worthless vagabond” when he was admitted for an “inflamed knee” in 1839. Alms House personnel made comments if they found a man to be temperate as well. Thus, Louis Fougeray, an insanity patient, was described as “temperate, studious, [and] melancholy” when admitted to the Alms House in August 1841. Likewise, a clerk commented that James Douglas, another insanity patient, was “said to be temperate” in February 1839. Apparently, the vast majority of those admitted to the Alms House were intoxicated to the extent that even a rumor that a patient was temperate was worth writing down. In this context, Alms House personnel listed intemperance as a “cause of poverty” primarily to designate men who were notably intoxicated at the time of admission.[5]
Mania a potu was also considered an insanity disorder. Medical writers often noted the analogy between the agitation of mania a potu patients and that of simple mania. Indeed, the similarities between mania a potu and mania were the main reason that doctors preferred mania a potu to terms like delirium tremens. Before 1835, both mania a potu patients and insane patients were placed in the cells so that they could be more closely monitored and controlled if they became agitated, violent, or attempted to harm themselves. After 1835, mania a potu patients and the insane were listed as being placed in the cells or the Lunatic Asylum. As was the case with alcohol, patients who suffered frequently from mania a potu could be classified as insane as well. That was the case with John P. Kelly, a printer who frequented the Alms House throughout the period of study. On April 1, 1837, Kelly was admitted for being “insane with drink” and placed in the “cells.” Kelly was back in the cells twice in 1838, but categorized on these occasions as “insane.” In 1840, Kelly was listed as “deranged” in March and as having mania a potu on his admission in June. [6]
Another area of considerable overlap involved transfers to the cells or Lunatic Asylum. A number of mania a potu cases developed after men were admitted to the Alms House. On April 30, 1838, for example, William Lucas, a peddler, was admitted to the alms house because he was destitute, but then developed a case of mania a potu and was transferred to the cells for treatment. Ten days after Lucas received his doctor’s release on May 13, 1838, a plasterer named Samuel Link was admitted to the Alms House for destitution and then developed his own case of mania a potu before being released three days later. Both men must have been extremely heavy drinkers who could not have obtained enough alcohol at the Alms House. As a result, they developed mania a potu symptoms and were transferred. Beginning in 1836, significant numbers of men were transferred from medical wards to the Lunatic Asylum (0 in 1835, 7 in1836, 33 in 1840). The likelihood of those cases involving mania a potu was especially high when men were listed as intemperate or intoxicated and then re-assigned to the Lunatic Asylum, but it is also likely that mania a potu was involved in the many cases where men were re-assigned to the Lunatic Asylum but no disorder was identified. Although it may be true that the anxieties attendant on confinement in the Alms House led some men to develop symptoms of mania or depression, it was much more likely that re-assignments to the Lunatic Asylum occurred because deprivation of alcohol had led to mania a potu.[7]
The other sources for data on mania a potu at the Philadelphia Alms House were two medical articles. These include an anonymous 1830 article by a doctor who supervised the “cells” at the Alms House and an 1843 journal article on deaths at the Alms House by John Prosser Tabb, a physician who did research at the Alms House for his medical dissertation. Both articles find many times the number of cases and deaths from mania a potu than are listed in the Men’s Register. For example, the anonymous doctor listed 70 cases and 9 deaths from mania a potu between November 1, 1828 and Feb. 1, 1829. During the same period, the Men’s Register listed no cases and no deaths under the term mania a potu or any synonyms for mania a potu like delirium tremens or the horrors. Likewise, the same author listed 75 cases and 18 deaths from mania a potu between June 10, 1829 and Sept. 10, 1829 where the Men’s Register only listed 7 cases and no deaths and the Women’s Register only recorded 9 cases and no deaths from mania a potu for the whole year of 1829. Given that the Register of Males and Register of Females had listings for all (or practically all) those admitted to the Alms House, then the Registers were not listing mania a potu for at least 59 of the men and women (78.6%) who were suffering from mania a potu in the Alms House during those three months.[8]
Similar discrepancies can be seen between the Alms House records and John Prosser Tabb’s article on deaths at the Alms House. Tabb derived his statistics on deaths from mania a potu between 1832 and 1843 from the death certificates issued by the Philadelphia Alms House.
Table 1. Tabb’s Statistics Men’s and Women’s Register
Mania a potu deaths Deaths Listed as Mania a Potu
1832 75 0
1833 43 15
1834 60 17
1835 19 7
1836 24 5
1837 43 4
1838 16 1
1839 13 2
1840 4 2
1841 3 1
1842 5 0
1843 8 4

Tabb found that death from mania a potu gradually declined between 1832 and 1841 before rising slightly by the end of the depression in 1843. Tabb also cites Alms House doctors as claiming that there was a 10% death rate from mania a potu between 1832 and 1837. Given a 10% death rate, 75 deaths from mania a potu in 1832 would have meant that there were approximately 750 cases of mania a potu at the Alms House during that year. That would be an average of 14.5 mania a potu cases a week. Nevertheless, the Men’s Register and Women’s Register only list 22 cases of mania a potu altogether for 1832 and no deaths. Although the discrepancies for the rest of the period covered by Tabb’s article are not as severe as those in 1832, they are still quite significant. In 1833-34 and 1836-39, the number of deaths recorded by Tabb were more than three times those recorded in the Men’s and Women’s Registers. In the other four years, the number of deaths recorded by Tabb was still at least twice the number listed in the registers. [9]
Why did the patient registers list so many fewer cases of mania a potu than were observed by the doctor directing the cells or recorded in the death records examined by Tabb? One source for the discrepancy was the diverse ways in which alcohol cases were recorded between 1828 and 1850. Changes in the personnel recording cases in the Men’s Register resulted in striking alterations in the recording of alcohol cases in 1832, 1834, 1837, and late 1848. In 1831, for example, there were 331 mentions of alcohol under three kinds of headings—16 mania a potu cases, 222 cases of intoxication as a medical condition, and 58 listings of intemperance as a “cause of poverty.” In Nov. 1831, however, a new clerk began recording entries into the Register of Males who seemed skeptical of alcohol-related diagnoses. Listings for mania a potu increased modestly from 16 to 22 in 1832, but the number of “intoxication” diagnoses dropped from 222 in 1831 to 50 in 1832 and 57 in 1833. Likewise, the intemperance as cause of poverty listing dropped from 58 in 1831 to 0 in 1832 and 8 before a new clerk took over in November, 1833. Where previous and subsequent clerks listed alcohol cases, the clerk for 1832-1833 used generic listings like “sick” or “eloped.” This is the primary reason why there were only 76 alcohol-related listings in the Register of Males for the whole year of 1832 even though Tabb listed 75 deaths from mania a potu for that year.[10]
The new clerk who began in November, 1833 took the other extreme. This clerk’s listings for mania a potu and intoxication were not radically different from other clerks. There were 51 cases of mania a potu (less than 1833) and 130 intoxication cases recorded. Nevertheless, the 1834 clerk listed 379 men as being intemperate as a cause of poverty. As a result, there were 560 listings of alcohol-related cases altogether for males. There were also 200 alcohol-related cases listed for women or 760 altogether. These numbers bear close examination. First, the 1834 register has about 160 more alcohol listings than the number of mania a potu cases that could be estimated from Tabb’s statistics. Tabb lists 60 men and women as dying from mania a potu in 1834. With a 10% death rate, there would have been around 600 cases of mania a potu at the Alms House compared to the 760 alcohol listings. Although the number of alcohol listings overshoots the estimated number of 600 mania a potu cases from Tabb, it is still much closer than the 51 cases listed in the Men’s and Women’s Registers. Second, it is most likely that the almost 550 cases of mania a potu that were not listed in the Male and Female Register would have developed out of the pool of 760 cases where patients were listed as intoxicated or intemperate. This is especially the case where patients were admitted for intoxication, but also would have been true of those who were listed as having intemperance as a cause of poverty. In these cases, the patient was either noticeably intoxicated upon admittance or had broken down physically as a result of alcohol consumption. It was out of this larger pool of noticeably intemperate patients that most mania a potu cases most likely would have arisen.[11]
Alms House records and medical articles also provide a basis for prioritizing the likelihood that cases out of the overall pool of alcohol-related listing would develop into mania a potu. Of course, the cases that were most likely to be mania a potu were those cases listed as mania a potu, delirium tremens, or some other synonym for mania a potu. There is also considerable likelihood that cases where patients were listed as intoxicated” and involved mania a potu. In compiling tables on mania a potu cases for his article, John Prosser Tabb listed mania a potu cases on a continuum of “intoxication, “delirium tremens—1st stage, delirium tremens—2nd stage, delirium tremens, 3rd stage.” As a result of his consultations with Alms House doctors, Tabb seems to have concluded that “intoxication” cases manifested enough symptoms of delirium tremens and mania a potu that they should be categorized as mania a potu and treated with either opium or alcohol. Cases where patients were listed as “intoxicated” and assigned to the “cells” or Lunatic Asylum were particularly likely to be mania a potu cases. Mania a potu cases were difficult to manage because of patient violence and almost all mania a potu patients were assigned to the cells. If a patient were recorded as “intoxicated” and assigned to the cells, he would be even more highly likely to be suffering from mania a potu than an intoxicated patient assigned to the medical ward or long ward. As a result, cases where the patient was listed as intoxicated and assigned to the cells were categorized as “highly likely” to be mania a potu cases along with those listed as mania a potu. Patients whose medical condition was listed as intoxicated and assigned to the medical ward still had a good likelihood of having or developing mania a potu while those patients who had “intemperance” listed as a cause of poverty should be viewed as having some likelihood of developing mania a potu. The typology below prioritizes the likelihood of alcohol-related cases being mania a potu from the highly likely Type I to the somewhat likely Type III.[12]
Type I—High likelihood of mania a potu
Mania a potu, delirium tremens, etc.
Intoxicated (cells or Lunatic Asylum)

Type II—Good likelihood of mania a potu
Intoxicated (medical ward or long ward)

Type III—Some likelihood of mania a potu
Intemperance listed as cause of poverty


Mania a potu also was subject to variable diagnosis. When it ran its full course, mania a potu cases had three stages. Patients often fell into mania a potu because gastric disorders (fevers, gastritis, blockages, etc.) made it impossible for them to drink their usual quota of alcohol. In the first stage, patients trembled all over their bodies and evidenced enormous fear concerning their environment. According to Willis M. Lea, first-stage mania a potu patients initially had a “countenance commonly expressive of great fear and anxiety or wild and starring though sometimes fixed and sullen.” These symptoms could then balloon into a second stage of fearsome hallucinations. At this stage, mania a potu patients had bursts of agitation and manic energy as they strived to hold up the walls in their rooms, defend themselves against imaginary attackers, or plead their innocence to medical personnel. If mania a potu was not cured in the second stage, barely-conscious patients fell into a final stage of convulsions and apoplexy that could exhaust their physical systems and result in death.[13]
The only point where mania a potu could be clearly distinguished was in the second stage when a patient was subject to hallucinations. The first stage of mania a potu was especially subject to variable diagnosis. Even though patients had not consumed alcohol for a period of time, they still had a considerable amount of alcohol in their systems. This is why men who suffered from mania a potu could so easily be diagnosed as “intoxicated.” Likewise, doctors could base their diagnosis on a patient’s initial trembling, fevers, chills, and/or constipation and mistakenly categorize mania a potu patients as sick, fevered, or having a stomach disorder. In such cases of mistaken diagnosis, the only indication of mania a potu in the Register of males would have been the listing of “intemperance” as a cause of poverty.
The end stages of mania a potu also presented difficulties in diagnosis. If a patient brought in during the third stage of mania a potu was unable to communicate the progression of the disorder, a doctor or clerk could mark down the illness initially as “apoplexy,” “convulsions,” or “effusion of the brain” even though these conditions had developed as a result of mania a potu. Once again, the only indication in the patient registers that such a case might have been mania a potu would have been a listing of intemperance as a cause of poverty.
There is also a strong likelihood that many mania a potu cases were listed under generic categories like “sick” or “eloped.” As noted before, these listings were used especially extensively in 1832 and 1833, and, as a result, there were few listings of either mania a potu or “intoxication” for those years. Many cases of mania a potu were hidden behind generic listings for insanity as well. Most insanity cases were listed under generic categories like “insane” or “deranged” and there was almost no effort to specify mania, monomania, hypochondria, hysteria, or other contemporary categories of mental disorder. Indeed, clerks often satisfied themselves by marking patients as being placed in the insane asylum and not listing a condition at all. Thus, although mania a potu was specified more often than other mental disorders, it is still very likely that many mania a potu cases were either listed under the generic categories of “deranged” or simply listed as assigned to the lunatic asylum or cells.[14]
As was the case in relation to alcohol disorders, the Registers of Males and Females also provide a basis for prioritizing the likelihood that a case listed as insane or deranged would be mania a potu. The strongest likelihood for mania a potu were those men who were listed as being admitted to the Alms House for an illness or injury and then were transferred to the lunatic asylum. It is highly likely that the heavy drinkers among the new residents developed mania a potu symptoms as a result of being denied alcohol on admission and then were transferred to the lunatic asylum. It is also possible to view those insane patients who had very short stays as moderately likely to have been suffering from mania a potu. Although the medical literature cites cases where mania a pot resulted in permanent insanity, mania a potu was generally a short-term disorder. This is especially the case in the late 1830’s and 1840’s as treatment improved. The percentage of mania a potu cases in which male patients were discharged by the doctors within a week increased from 38.5% in 1829 to 76.3% in 1842, and 80% in 1846. Thus, there is a good likelihood that men listed under generic insanity categories like “insane” or “deranged” and released within a week suffered from mania a potu.[15]
There are two further considerations which make this conclusion plausible. The first concerns the relative frequency of short-term insanity cases among men and women. Although short-term insanity cases of less than a week of duration were relatively frequent among males, very few female patients spent a week or less being treated for insanity. It is most likely that women were less susceptible to short-term insanity because of their lower alcohol consumption. Not only did women in the Alms House have fewer mania a potu cases, but they also had far fewer intoxication cases or cases where intemperance was listed as a cause of poverty. Moreover, there were almost no cases where women were shifted from medical wards to the lunatic asylum. In this sense, the low number of short-term female insanity cases was correlated with the relatively low number of alcohol disorders among women. Conversely, the relatively high level of short-term insanity among males was associated with the pervasiveness of male alcohol disorders. Because many, perhaps most, short-term male insanity cases at the Philadelphia Alms House involved mania a potu, these kinds of cases have been categorized as Type II cases having a “good chance” of being mania a potu cases.[16]
There is also some likelihood that cases of insanity that stretched out as long as a month were also mania a potu cases. In the late 1820’s and early 1830’s, about one in four mania a potu cases (27% in 1829, 22.5% in 1830) lasted between one week and one month. The proportion of these kinds of cases declined as medical care for mania a potu improved during the late 1830’s and 1840’s. However, it still took one in five mania a potu patients more than a week and less than a month to be released from the Lunatic Asylum in 1842 (19.9%) and 1846 (18.7). Given that there would have been a substantial number of mania a potu patients among those listed as insane or deranged, it is likely that some of those who were treated between one week and a month for insanity were mania a potu patients. Although cases in this group would not have been nearly as likely to have mania a potu as short-term insanity patients, they would have included enough mania a potu patients to be listed as Type III, “somewhat likely” to be mania a potu patients.[17]
The complete typologies for the relative likelihood of mania a potu cases among the listings for alcohol disorders and insanity at the Philadelphia Alms House are listed below. Almost every one of the Type 1 cases would have been mania a potu and most of the Type II cases as well. Intoxication cases where patients were placed in the medical or long wards were listed as Type II because Dr. Tabb categorized such cases as forms of mania a potu. There also was a good chance that very short-term insanity cases (a week or less) would have entailed mania a potu as well. To the contrary, there is only some likelihood that Type III listings would have been mania a potu cases.
Type I—High likelihood of mania a potu
Mania a potu, delirium tremens, etc.
Intoxicated (cells or Lunatic Asylum)
Intemperance, Insanity
Transfers to Lunatic Asylum

Type II—Good likelihood of mania a potu
Intoxicated (medical ward or long ward)
Short-term insanity cases (a week or less)

Type III—Some likelihood of mania a potu
Intemperance listed as cause of poverty
Medium-term insanity cases (between a week and a month)

3. Testing mania a potu scenarios.
In the final section, several scenarios concerning mania a potu cases will be tested in the effort to make the estimate of male mania a potu cases from the Register of Males approximate to the findings of the ante-bellum medical literature more closely. In Graph 1 below, the number of deaths listed in Tabb’s Statistics of Death (blue line) is compared to the listings for deaths from mania a potu (pink line) in the Men’s and Women’s Register and the more general Type I listings (yellow line). As already mentioned, relatively few men and women who were listed as having mania a potu died at the Philadelphia Alms House. While Tabb recorded highs of 75 deaths in 1832, 60 in 1834 and 43 in 1837, the highest number recorded in the registers was 17 for 1834. The only time when the lines representing Tabb’s statistics and the mania a potu listings come close to intersecting is between 1840 and 1843 when the number of deaths in Tabb came close to zero. Even in those years of low death numbers, Tabb listed double or more the number of deaths listed as mania a potu cases in the Registers. Tabb recorded 20 deaths for 1840-1842 while there were only 7 deaths listed as mania a potu cases.
Including all of the Type 1 listings as mania a potu cases almost doubles the number of deaths from those listed as mania a potu cases for 1832-1835 (from 39 to 71), but that increase does not greatly close the gap between the deaths recorded by Tabb and those in the patient registers for those years. One of the most important reasons why the number of mania a potu deaths did not increase that much when all Type I cases were counted was that the death rate for those listed as mania a potu cases (19.44%) was almost twice that as Type I cases as a whole (10.58%). In this context, it is likely that the Alms House clerks of the early and mid-1830’s only listed the most obviously hallucinating of mania a potu cases as having mania a potu. As a result, counting the other Type I cases as mania a potu brings the death rate for mania a potu cases more in line with the 10% death rate recorded by Tabb, but did not greatly decrease the gap between Tabb’s statistics of death and listings in the Register of Males and Register of Females.[18]
Counting the Type I cases as mania a potu does close the gap between Tabb’s statistics and the listings in the Register of Males and Register of Females between 1840 and 1843. In 1840 and 1841, there were 7 deaths listed in Tabb and 8 in the Type I cases listings. The gap began to open back up in 1842 (5 in Tabb, 2 Type I) and 1843 (8 in Tabb, 5 in Type I). Altogether, Tabb listed 20 deaths for 1840-1843 while there were 15 Type I listings. Of course, part of the reason that the gap closed for 1840-1843 was that it was difficult to fall very far below the low number of deaths listed by Tabb. At the same time, improvements in the recording of mania a potu cases brought the Type I listings more in line with Tabb’s statistics. The recording of mania a potu cases had begun to improve in 1836 when Alms House personnel began to list significant numbers of cases that had changed to mania a potu after admission. By 1839, Alms House personnel were recording more changed diagnoses and transfers to the Lunatic Asylum and the number of Type I cases leaped from an average of 99.8 from 1834-1838 to 201.6 from 1839-1843. One reason why Type I listings of deaths from mania a potu were closer to Tabb’s statistics for 1840-1843 was improved record keeping.[19]
In Graph 2, Tabb’s statistics for mania a potu deaths (blue line) are compared to the death totals from Type I cases, Type I-II, and Types I-II-III. For the early and mid-1830’s, counting Types I, II, and III (red line) as mania a potu cases is most advantageous. Considering Type I, II, and III together accounts for almost 2/3rds (38) of the 60 mania a potu deaths that Tabb finds in 1834 while taking into account Type I and II (green line) only accounts for 1/3 of Tabb’s mania a potu deaths (20 of 60). Given that Alms House personnel did a poor job of listing incoming mania a potu cases and failed to record cases of mania a potu that developed within the Alms House, there would have been many mania a potu cases hidden within the “intemperate” listings for those years. For the late 1830’s, however, Type I-II-III and Type I-II track equally well with Tabb’s statistics. As the recording of mania a potu improved later in the decade, the number of mania a potu deaths in the Type I-II listings began to better approximate Tabb’s findings. For 1835-1836, 1838-18-39, the number of Type I-II deaths was 17% less than those found in Tabb (60 in Type I-II to 72 in Tabb). At the same time, the number of Type I-II-III listings overshot Tabb’s statistics by 18% (72 in Tabb to 85 in Type I, II, III). Because Alms House clerks were reporting a higher percentage of mania a potu cases in Type I categories, fewer mania a potu cases were being listed just as “intemperate.” As a result, more of the deaths listed in Type III would not have been from mania a potu and counting Type III cases as mania a potu began to result in overshooting Tabb’s findings. Thus, the Type I-II and Type I-II-III listings for mania a potu deaths would both be useful for studying the late 1830’s.[20]
Historians researching mania a potu at the Philadelphia Alms House during the 1830’s would be wise to employ both the Type I-II and the Type I-II-III estimates. Because both methods produce estimates of mania a potu deaths that are close to Tabb’s findings for the late 1830’s, both estimates would be useful for studies of the individual medical histories of mania a potu patients, death rates from mania a potu at the Alms House, and the distribution of mania a potu by race, ethnicity, gender, and trade. Having access to two (relatively) valid methods of estimating the number of cases and deaths introduces a fair amount of flexibility into research on mania a potu. It makes it possible to confirm or disconfirm the findings gained by one method. For example, if one determined that the average age of the Type I-II cases that had mania a potu was 35, confidence in that finding would be increased if the average age of Type I-II-III cases was the same.
In much the same way, Type I and Type I-II estimates can be seen as effective ways to account for mania a potu cases and deaths in the 1840’s. Where considering all three types of cases as mania a potu results in overshooting Tabb’s statistics for 1840-1843 by 150% (50 to 20), considering the Type I and Type II cases results in less of an overshoot (37 to 20). In fact, the Type I-II-IIII estimate overshoots so much that it is more useful to examine the Type I-II estimates in relation to Type I estimates. Partly because of the low numbers and partly because clerks were doing a better job of listing mania a potu cases, the Type I estimates for mania a potu deaths from 1840-1841 are almost exactly the same as Tabb’s while the Type I-II estimates overshoot significantly.
Mania a Potu Deaths at the Philadelphia Alms House
Tabb’s statistics Type I Type I-II
1840 4 4 11
1841 3 4 10
1842 5 2 4
1843 8 5 12[21]

Moreover, as the deaths found by Tabb begin to rise again, the Type I estimates of mania a potu deaths at the Alms House remain as close to Tabb’s findings as the Type I-II estimates. When the blue line in Graph 2 for Tabb’s statistics of death stops in 1843, it is between the lines for the Type I and Type I-II estimates. Given that the methods for recording mania a potu cases at the Alms House remained consistent through 1847 (when record-keeping deteriorated), it is likely that the number of deaths from mania a potu remained somewhere between the Type I and Type II estimates until 1848. As was the case with Type I-II and Type I-II-III estimates for the 1830’s, it would be best to employ both Type I and Type I-II cases for studies of mania a potu during the 1840’s. In most cases, the data on cases identified by the Type I method would be somewhat more conservative than the data identified by the Type I-II method, but that can be factored into any analysis.
One example of how the Type I, Type I-II, and Type I-II-III methods of identifying could be considered together would be determining the death rate of mania a potu patients. The death rates identified by all three methods between 1829 and 1850 are shown in Graph 3 below.[22]
All three methods show a trend in which deaths rates from mania a potu decreasing from over 10% in the early 1830’s to below 5% in the early 1840’s and then rising slightly. As a result, there can be a great deal of confidence that death rates from mania a potu were lowered dramatically in the late 1830’s and early 1840’s. The existence of three methods also makes it possible to be flexible in approaching the data. For example, the estimates of death rates based on Type I-II cases appear to be most valid for 1829 and the early 1830’s. The Type I estimates (pink) would not be very significant in the early 1830’s because of the small number of Type I cases. Likewise, the Type I-II estimate (yellow curve) of the death rate for 1829 (18.28) fits much better with the death rate given in the article from the anonymous doctor (19.4%) than the estimate derived from the Type I-II-III method (11.99). The decrease in the death rates estimated by the Type I-II-III method in 1834 and 1835 is also suspect because of the excess in the “intemperance as a cause of poverty” listings for those years. For the 1840’s, the Type I-II and Type I-II-III methods derive similar results until about 1848 and provide some reinforcement for each other. There is a small but significant difference between the Type I-II and Type I approach (1.5% per year). If a study of death rates were being carried through the 1830’s and 1840’s, it might be better to employ the Type I-II approach for the sake of consistency. If one preferred a more conservative approach to ensure against overstating their argument, the Type I data might be advantageous because there is a higher likelihood that the Type I cases were mania a potu. Those seeking to capture as many mania a potu cases as possible might prefer to cast a broader net with the Type II approach. It would also be possible to average the results of the Type I and Type I-II approaches for a combined result or view the death rate in terms of ranging between the Type I and Type I-II outcomes.[23]
Conclusion.
The Alms House is the largest and most important of the many sources of data on mania a potu ante-bellum Philadelphia. Unfortunately, however, relatively few of the mania a potu cases at the Alms House were recorded as mania a potu or one of its synonyms. Consequently, the medical data at the Philadelphia Alms House is only useful for historical research if there is some way of determining which cases were likely to be mania a potu even though they were not listed as such. To address this difficulty, three different approaches to defining mania a potu cases (Type I, Type I-II, and Type III) within the Alms House records were defined and compared to the data on mania a potu deaths in John Prosser Tabb’s “Statistics of Death at the Philadelphia Alms House.” Because of the continually changing character of Alms House record keeping, all three methods are more applicable for some years between 1829 and 1848 than others. As a result, it is possible to use each of the three different approaches in a flexible way to double-check, reinforce, qualify, or reject the results using the other approaches. Thus, the Alms House records, though still not completely reliable, provide a large, flexible, and very useful data base for the study of mania a potu in Philadelphia.



[1]Herman Melville, White Jacket, Northwestern University Press, 1970, 242
[2]John Prosser Tabb, Tabb’s Statistics of Deaths in the Philadelphia Hospital,” American
Journal of the Medical Sciences, 1844.
[3]Register of Males and Register of Females in the Alms House Records, Philadelphia City Archives. Both registers were divided into books spanning 1828-1836, 1836-1843, 1843-1847, and 1847-1851. The Register of Males was consulted for 1828-1850 and the Register of Females for 1828-1843.
[4]For the beginning of placement in the Lunatic Asylum, see listings in the Register of Males for 1834 and 1835. Alexander Draper was a well-known labor and temperance speaker who was admitted to the Alms House 106 times between 1834 and 1842. For 1838 Alms House visits when Draper was diagnosed with mania a potu, see Register of Males, Feb. 3, April 29, May 29, June 15, and July 6; for stays at the Alms House when Draper was diagnosed as intoxicated, see Register of Males, July 27, Sept. 9, and Oct. 23. For Draper’s public speaking, see his speech on behalf of seamstresses, Pennsylvanian, June 26, 1835 and a temperance speech before the Howard Temperance Society on “intemperance and mania a potu,” Public Ledger, August 31, 1841.
[5]For Samuel Mulligan, see The Register of Males, March 30, 1841; for Samuel Porter, Ibid, Sept. 3, 1839; for Louis Fougeray, Ibid, Aug. 17, 1841; for James Douglas, Ibid, Feb. 13, 1839.
[6]For analogy with mania, see William Milnor, “An Inaugural Dissertation on Mania e’ Temulentia,” medical dissertation, University of Pennsylvania, 12-13; Isaac C. Snowden, “An Inaugural Essay on Mania a Potu,” 1817, 192; for John P. Kelly, see Register of Males, April 1, 1837, Jan. 27 and Mar. 19, 1838, March 10 and June 22, 1840;
[7]For William Lucas, Register of Males, April 30 and June 4, 1838; for Samuel Link, Ibid, May 23, 1838; the count for the number of men transferred to the Lunatic Asylum was compiled from the Register of Males, 1828-36, 1836-43, 1843-47.
[8]Anonymous, “Observations on Mania a Potu,” n.d. (probably 1830), Historical Society of Pennsylvania, 8.
[9]Tabb, “Statistics of Deaths in the Philadelphia Hospital,” 364; statistics for Alms House compiled from Register of Males, (1828-1836, 1836-1843,1843-1847) and Register of Females(1828-1836,1836-1843), Philadelphia Alms House, City Archives of Philadelphia.
[10]Totals for mania a potu, intemperance, and “alcohol as cause of poverty” listings for 1831, 1832, and 1833 compiled from the Register of Males, 1831-1833.
[11]Totals for alcohol listings for 1834 were compiled from the Register of Males, 1834.
[12]Tabb, “Statistics of Deaths in the Philadelphia Hospital,” 365; for one of many cases in which a patient diagnosed as “intoxicated” was assigned to the Lunatic Asylum, see the listing for James Caulfield, Register of Males, August 22, 1838.
[13]For the connection between gastric disorders and mania a potu and first stage of the disorder, see Willis M. Lea, “An Essay on Mania a Potu,” 3;
[14]For examples of a patient listed only as assigned to the lunatic asylum, see Register of Males, entry for Samuel Primrose, July 16, 1838 and William Eberton, May 5, 1843.
[15]The percentages for mania a potu patients discharged within a week were compiled from the Register of Males, 1828-1836, 1836-1843, 1843-1847.
[16]For female patients, see the Register of Women, 1828-1836, 1836-1843. For an example of the extent to which there were fewer female mania a potu cases, there were 1038 male Type I and Type II mania a potu cases from 1840-1843 and 228 female Type I and Type II cases during the same period (compiled from Register of Males and Register of Females).
[17]Percentages for mania a potu cases that lasted betw een a week and a month were determined from the Register of Males, 1828-1836, 1836-1843, and 1843-1847.
[18]Graph 1 was compiled from Tabb, “Statistics of Death in the Philadelphia Hospital,” 364, Register of Males, 1828-1836, 1836-1843, and 1843-1847, and Register of Females, 1828-1836, 1836-1843, and 1843-1847. Like Tabb’s statistics of mania a potu deaths, the estimates of mania a potu from the Alms House registers incorporate data for both males and females.
[19]For counts for Type I cases from 1834-38 and 1839-43 were compiled from the Register of Males and Register of Females, Alms House, 1828-1836, 1836-1843, and 1843-1847.
[20]The statistics for Type I-II and Type I-II-III estimates were compiled from the Register of Males and Register of Females, Alms House, 1828-1836, 1836-1843, 1843-1847, and 1847-1851.
[21]Table compiled from Tabb, “Statistics of Death at the Philadelphia Hospital,” 364 and Philadelphia Alms House, Register of Males (1836-1843,1843-1847) and Register of Females (1836-1843).
[22]Graph 3 was compiled from Anonymous, “Observations on Mania a Potu,” 8; Philadelphia Alms House, Register of Males, (1828-1836, 1836-1843,1843-1847) and Register of Females (1828-1836,1836-1843).
[23]Philadelphia Alms House, Register of Males, (1828-1836, 1836-1843, 1843-1847) and Register of Females (1828-1836,1836-1843).