Burke, Horwood and White's floor plan of the house, however, reveals little integration between the front elevation and the spaces inside, which were much more commercial and scientific than the homey exterior suggested. Behind the bay window, to the right of the front door, was the physician's office; across the hall, boasting white pine trim, oak borders, and a gas fireplace, was a waiting-room for patients. Indeed, nearly half of the ground floor area of the Dumble house was given over to his medical practice, with a close and direct connection to rooms presumably used by his family and live-in servant(s): the dining-room, stairway, and kitchen. A number of initial questions flow from the Dumble house example and frame our general interests in this combination of family and work: which physicians invested in home practices and why? How did the home-office function? What was its relationship to the traditional family home?

This interdisciplinary paper explores the design and distribution of home-offices or house-offices³ in Toronto, Ontario, using data gathered from city directories, nominal census returns, architectural drawings, photographs, obituaries, and death notices. Our approach combines interests in architectural analysis and the historic role of physicians in the design of everyday medical space with a demographic and social perspective on the nature and distribution of specialized households in an urban context.⁴ The approach is cross-sectional, with a direct and comprehensive examination of physicians practising in 1901 Toronto, which, at that time, was Canada's second largest city with a population of almost 240,000.⁵ The genesis of this work was the discovery of a selection of architects’ drawings for home-offices, including Dr Dumble's, in the Archives of Ontario.

The paper is structured around three main sub-arguments. The first challenges the common perception that urban physicians practised medicine in more sharply-defined and formalized spaces than their rural counterparts.⁶ Of the 342 physicians advertising their practice through listings in Might's Toronto City Directory in 1901, 311 or 91 per cent maintained a house-office. As there were a number of other physicians practising in Toronto at this time who did not advertise in the directory, our study focuses only on those actively promoting their independent practices.

Interior photographs of the offices of Francis Buller, an oculist, and Herbert Stanley Birkett, a laryngologist, in 1890 Montreal reveal the ambiguous and informal feel of the urban physician's home-office (Figure 2). The presence of the safe, interior finishes and general ambience in photographs of Buller and Birkett's premises, for example, provide material evidence of the profitability of urban medicine at this time. The conspicuous display of books (symbolizing medical knowledge), medical instruments, fine furniture, and carpets, are signs of prestige, alluding to the physicians’ credentials and status. These domestic trappings were a significant aspect of what the medical historian Edward Shorter has defined as the “modern” doctor–patient relationship in the pre-Second World War period, framed by “traditional” and “postmodern” periods. According to Shorter, the physician's authority rested on the appearance (his emphasis) of caring, expressed in actions such as the laying on of hands, counselling patients in intimate problems, and privileging personal history.⁷ The domestic setting contributed to this image of the doctor as an understanding confidant, almost like a member of the family, while the conspicuously expensive touches reinforced the cultural and material choices of fee-paying patients. As George Bernard Shaw said in his preface to the 1911 The doctor's dilemma, “[Doctors] must believe, on the whole, what their patients believe, just as they must wear the sort of hat their patients wear”.⁸ Like Shaw's hat, the home-based office, in its subtle overlap with the family home, subconsciously linked patients and their physicians. The floor plans of home-offices, too, show how glimpses of the doctor's family life may have boosted his or her professional roles.

This model of the physician home-office emerged from larger structural and technological changes that were occurring in Canada (and, more broadly, North America) at the time. The standard medical model of practice had traditionally been associated with physicians attending patients in their own homes, in a home-visiting model of practice. As Neil Larry Shumsky, James Bohland, and Paul Knox have argued, “the lack of urban transportation meant that the basic setting for health care was the patient's bedside rather than the doctor's office”.⁹ The introduction of mass transit in urban centres (such as the streetcar) offered an important opportunity to modify this model of care, as patients could now reliably travel to a physician's office. Further structure was instilled with the advent of the telephone which, according to Paul Starr, “made it much easier for patients to see physicians at their offices at a prearranged time, reducing the risk of dropping in while the doctor was out on call; it also made office practice more attractive to doctors, who could now make orderly appointment hours and see more patients than when relying on an uneven stream, or trickle, to their door”.¹⁰ The net result was an increased efficiency in health care delivery, at least from the perspective of physicians, who would appreciate the notable reduction in the portion of their work-day wasted on travel. Knox, Bohland and Shumsky argue that in America rising physicians’ fees “must have been an incentive for people to substitute their own time for that of the doctor by travelling to his office instead of paying him an extra fee to visit them at home”.¹¹ In Toronto, the first horse-drawn streetcars and sleighs (winter) were introduced in 1861 and the first electric streetcars appeared by 1892,¹² both enhancing the potential success of the home-office. Knox, Bohland and Shumsky further argue that the transition of health care away from family-centred care to science- and technology-driven approaches, with associated specialized equipment and a greater reliance on paramedical assistants and clerical support, meant that physicians needed their own space, a dedicated medical office space, to house this new approach.¹³

The era of home visits overlapped with the advent of home-offices. Sometimes physicians would continue to practise both models of patient care, as circumstances demanded. One of the physicians in our study, Dr Herbert J Hamilton, began his Toronto practice in 1894 and commissioned a purpose-built home-clinic in 1906, the same year that he became an elected member of the senate of the University of Toronto. Hamilton died in 1920 (aged fifty-five), reportedly falling victim to influenza towards the end of the Spanish flu pandemic.¹⁴ He had been sick for some three weeks before his death, but carried on with his duties and died shortly after collapsing in the home of a patient he was treating for the flu. This detail of his last days, captured in a newspaper account of his death, highlights the fact that even urban physicians who had committed to the home-office model of medical practice would still home-visit, particularly if patients were too ill (or incapacitated) to travel to the physician's office.

Secondly, we argue that the location of the house-offices foreshadowed and perhaps facilitated the development of Canada's most coherent hospital district. Harley Street in London¹⁵ and the Longwood Medical and Academic Area in Boston are examples of the same sort of area in Britain and the United States. For scholars who study vernacular architecture, this relationship between the house-office and the city is an important reminder that less visible, hybrid, non-monumental buildings may be as significant or interesting as monuments, in this case hospitals.¹⁶ Toronto's 1892 Victoria Hospital for Sick Children (later called the Hospital for Sick Children) (Figure 3), located at the heart of the cluster of the house-offices explored in this paper, was a central node for professional and social networking among medical professionals in turn-of-the-century and today's Toronto.¹⁷ Significantly, this historic postcard of the institution includes two streetcars. We speculate that if physician home-offices were strategically placed to take advantage of easy access for patients, then the same advantages would accrue for the hospitals that came later. Many of Toronto's hospitals, particularly those for women and children, started off in private homes, as was the case in cities across North America, another significant link between houses and hospitals.¹⁸

Functional zoning was another strategy employed to separate family and patients. In Dr Herbert J Hamilton's house (Figure 6) of 1906 the architect's strategy was to separate the domestic and professional zones to the left and right of the entry hall. In all cases, despite variations in the plan configurations, the medical function of the house was completely disguised on the exterior of these structures.

Dick drew up two sets of plans in configuring the house-office of Dr Henry Machell²⁸ (Figure 9), experimenting with connecting the surgery and waiting-room separately from the family space. Like the Graham plan, the Machell household could be effectively isolated from patients simply by closing an inner vestibule door. The doctor's practice occupied the left side of the central hall, while the family spaces were on the right. A narrow passageway, located behind the family's stair, linked the patients’ waiting-room with a surgery. A rather ingenious sequence of interior doorways presumably shut off the family space from the practice.

Dr Thorburn's house-office of 1891 (Figure 10), also by Dick, is one of the few in the sample to include an entirely independent side entrance for patients, allowing them to enter directly into the heart of the house with no overlap whatsoever with the more private family spaces. In this case the rooms given over to medical practice were a waiting-room, a consulting room and a small laboratory, each with a window. In 1901 the Thorburn household included the doctor, his wife Jane, and his adult unmarried daughter Georgina. Jane Roseman and Julia Malloy were employed and living in the home as a domestic and a cook respectively. Not surprisingly, Dick's design includes two servants’ bedrooms above the kitchen. Also in the family home was Ernest McPhee, thirteen years of age, who had emigrated from England in 1900. According to the census, this was just one of three dwelling houses owned by Thorburn, along with 600 acres of real estate.

It is important to note that even these early examples have generous waiting-rooms that occupy a substantial percentage of the medical practice area. Some waiting-rooms, such as Dr Dumble's, occupied nearly 25 per cent of the ground floor of his house; and many included finely finished fireplaces with mantles, like the one that shows in Dr Hamilton's plan. In Dr Eaton's house-clinic, the waiting-room is larger than either of the two operating rooms. The presence of these elaborate spaces for waiting may imply that the physicians anticipated seeing multiple, overlapping appointments in a single time period, reinforcing the idea that doctors were increasingly turning to a business-based, time-centred model and away from the more haphazard experience of being called to patients' homes. At the same time, these waiting-rooms also provided comfortable spaces for a patient's family or companions to wait during consultations.

It is likely that the distinct spatial separation observed in our earliest plans generally diminished over time, as there is a striking fluidity between family life and practice in later home plans. As a case in point, in the much later 1905 Eaton house plan, and other examples, patients and family members shared a common entry sequence, but patients were then diverted immediately into the spaces devoted to practice. Sometimes the separation might be as minimal as a glass door. This openness and uninterrupted flow of space marks a departure from earlier examples where there seemed to be a preference for maintaining separation between the patients and the household, either through a separate access, like the Thorburn house, or fitting new passageways under staircases, as in the Machell plan.

When doctors retro-fitted houses as house-offices, public rooms moved upstairs and downstairs in order to accommodate waiting-rooms, surgeries and laboratories. In quite a few cases billiard-rooms were re-located to the basements, and sitting-rooms moved upstairs, suggesting a vertical displacement of family life as the medical practice came to occupy most of the ground floor. It is impossible to know from the architectural evidence whether families used waiting- and/or medical rooms for their own activities when they were not being used by the physicians. In all cases, however, even when the medical practices came to take over the ground floor, the kitchen always remained relatively inaccessible from the rooms dedicated to practice.

In the case study showing the greatest fluidity, two physicians’ families lived side-by-side in semi-detached homes. In 1898, Dr George A Peters,²⁹ associate professor of clinical surgery (since 1890) at the University of Toronto, and Dr Alexander Primrose,³⁰ University of Toronto professor of anatomy (since 1896) and associate professor of surgery (since 1897), commissioned a three-storey duplex from Dick at 100–102 College Street, with consulting and waiting-rooms occupying the first two rooms on the exterior walls of each twelve-room dwelling (Figures 11–12). Note that in the Peters–Primrose plan consulting and waiting-rooms were entered from the same hall that family members would use to enter the drawing-room or to access the staircase to the more private rooms of the upper level, with no separation whatsoever. In a surprisingly open situation, the doctor's waiting-room occupies the heart of the family home with an extra wide doorway that looks onto the staircase. Again, only pocket doors separate the waiting-room from the hall and there is no direct connection between the waiting-room and the consulting room. Patients would thus enter the family hall twice, once to go into the waiting-room and then again to see the doctor in the consulting room.

Both Peters and Primrose were surgical specialists, a medical niche perhaps facilitated by their urban environment. Both men remained in Toronto following their medical training, though their early lives were spent in rural locales (Peters was born in Eramosa, Ontario; Primrose was born in Pictou, Nova Scotia). Peters and Primrose were not only neighbours and colleagues in surgery, but also friends, as Primrose was one of eight pallbearers at Peters’ funeral.³¹

The locations of the remaining 31 (9 per cent) Toronto doctors who advertised separate home and office addresses in the Directory were also mapped (Figure 14). In this map we distinguish between Group 2 doctors, the 28 doctors with separate home and office addresses, and Group 3 doctors, the remaining three who advertised home-office as well as separate office addresses. Using home address and office address, the average distance travelled by “commuting” physicians was approximately 1.5kilometres each way. Dr Herbert Galloway undertook the shortest commute, estimated at 2metres, between his residence at 14 Bloor Street East and his office at the Toronto Orthopaedic Hospital located at 12 Bloor Street East. Dr John T Duncan undertook one of the longer commutes, about 5kilometres each way, between his home at 32 Fuller Avenue (the more southerly of the two black dots located towards the bottom left-hand corner of the map) and 39 Bloor Street East (the black triangle located to the immediate right of No. 27 the Toronto Orthopaedic Hospital).

Of the 244 physicians found, 95 per cent (n=232) and 5 per cent (n=12) were male and female physicians and surgeons respectively. Although Might's Directory has listings for only a handful of female physicians relative to male physicians, a number of notable female physicians were captured in these listings. Included are mother and daughter Dr Emily Stowe's and Dr Augusta Stowe Gullen's home-offices at 463 and 461 Spadina Avenue (close to the intersection with College Street) respectively. Though trained (by necessity) in the United States, Stowe is recognized as the first woman to practise medicine in Canada.³⁶ Not long after breaking a hip in 1893, Stowe retired from practice.³⁷ At the time of the 1901 census, then aged sixty-eight, she resided with her 38-year-old son Frank, a dentist, his wife Mary (aged thirty-four), their two daughters (Marie, eight, and Hilda, six) and son (Hudson, seven months), and Sarah MacLennoy, a 23-year-old domestic. It is puzzling that she continued actively to advertise a practice in 1901, even though her biographers suggest she had retired by this time. Stowe passed away only two years later, in 1903. Next door, at 461 Spadina Avenue, lived Stowe's daughter, Dr Augusta Stowe Gullen, also significant in Canadian history as the first woman to earn a Canadian medical degree (in 1883). Both Augusta and her husband, Dr John B Gullen, advertised their combined home-based practice. Living with them were Augusta's cousin (Emily Stowe) and Catharine Murphy, a domestic.

Women physicians lived in a range of familial and spatial situations. Emily Stowe was the only widowed female physician in the group of twelve women physicians advertising in Might's Directory. Along with Stowe Gullen there were three other married female physicians. Ida Lynd, forty-eight years of age, lived with her husband, a 53-year-old health inspector, their three children, two child “friends”, and a servant. Lynd reported earnings of $3000 a year, while her husband reported earnings of $600 a year. Margaret Gordon, forty-two years, lived with her husband, a forty-year-old salesman (earning $400 a year) and their two children. Jennie Coons, thirty, and her husband William Coons, twenty-four, and also a physician, boarded together in a home. The remaining seven female physicians were unmarried. Two lived in homes headed by their widowed mothers that included unmarried siblings. One woman lived in the home headed by her widowed father (no siblings). Julia Thomas, a 52-year-old single female physician, lodged in the home of a thirty-year-old piano maker and his family. Bertha Dymond, a 34-year-old surgeon, ran her own rooming house with a total of nine boarders recorded at the time of the census. Leila Davis, thirty-five years old and unmarried, was the only female physician living alone.

Returning to those physicians advertising home-offices in 1901, important findings emerge concerning ancestry and earnings. Overall, 99.2 per cent (n=242) of the home-office physicians self-identified as Canadian. Most of the physicians were born in Canada (84.9 per cent, n=207), with a small number coming from England (6.6 per cent), Scotland (2.5 per cent), Ireland (2 per cent), the United States (2 per cent), the West Indies (0.8 per cent), and “other” countries (1.2 per cent). 42.6 per cent of the physicians (n=104) claimed English/Welsh ancestry, while 25.4 per cent and 24.2 per cent claimed Scottish and Irish ancestry respectively. Of the remaining physicians, 6.5 per cent claimed other European ancestry and 1.2 per cent were unknown. The mean age of the physicians who advertised home-based practices was 45.3 years (sd=12.6), with the youngest physician reported as twenty-three years old and the oldest as eighty-five. Only 34 of the 244 physicians reported their earnings to the census enumerator. We calculated mean earnings of $2041.18 (sd=$1406.93) per annum among those who did report, with a minimum of $100 (Dr Thomas Phillips, aged sixty-six) and a maximum of $7000 (Dr James Cotton, aged thirty-nine).

Further insights can be gained into the nature and structure of home-office physician households from the 1901 nominal census returns. The mean household size was 5.3 (sd=2.1) persons. Physicians living alone occupied the smallest households, while the largest non-institutional physician household included fifteen persons. From our sample of 244 physicians, 218 (89.3 per cent) self-reported as the head of their households. Of those remaining, 4.1 per cent (n=10) were the son of the head, 1.2 per cent (n=3) were the daughter of the head, 1.2 per cent (n=3) were the wife of the head, 2.4 per cent (n=6) were “boarders and lodgers” and 1.6 per cent (n=4) were other family relations of the head (father, mother, brother-in-law, sister). Our results suggest that, in 1901, home ownership was not a prerequisite for establishing a home-office, since even a physician boarding or lodging in a household could presumably take on extra rooms to establish a home-based practice.

As the examples provided among woman doctors suggest, home-office physicians lived in a range of family and household structures. The majority of these physicians were married (80.7 per cent, n=197) at the time of the 1901 census, while a smaller percentage were unmarried (15.6 per cent, n=38) or were widowed (3.7 per cent, n=9). Home-office physician households were classified according to a truncated Statistics Canada household classification scheme (truncated because common-law unions were not recognized in 1901). Within this scheme there are nine possible household types: husband/wife (Type 1), husband/wife and additional persons (Type 2), husband/wife/never married children (Type 3), husband/wife/never married children and additional persons (Type 4), one parent/never married children (Type 5),³⁸ one parent/never married children and additional persons (Type 6), multiple family (Type 7), lone individual (Type 8), and a non-family grouping (Type 9). Figure 16 depicts the distribution of these household types in Toronto physician home-office households and in Canadian 1901 homes more generally (based on an analysis of the Canadian Family Project's 5 per cent sample of the 1901 population of Canada).³⁹ Relative to the average Canadian household, Toronto physicians and their families were significantly more likely to be living in households with “additional persons” (household types 2, 4, and 6). The nuclear family predominates, both for Toronto physicians as well as average 1901 Canadian households. But the important difference between the two is the overwhelming presence of “additional persons” living alongside physicians’ families. In the majority of 1901 Canadian households, 47 per cent of households, the nuclear family lived alone—mother, father, and unmarried offspring. For the majority of Toronto home-office physician households, however, 46 per cent were nuclear families living together with an assortment of “additional persons”. Though proportionately less common, the same tendency was true among husband/wife households (where Type 2 households with additional persons are more common among Toronto physicians) and single parent/offspring households (where Type 6 households with additional persons are again more common among the physicians). While 1901 households in general were much more open to extra-family persons relative to Canadian households today (which tend to be restricted to family members),⁴⁰ physicians’ families stand out in this regard.

A more detailed examination of the precise identities of these additional persons reveals a predominance of non-related persons. Among Type 2 (husband/wife and additional persons) physician households, for example, the majority of additional persons (77.2 per cent, n=44) were domestic workers (including maids/servants/nurse/page); of those remaining, 14.0 per cent (n=8) were relatives, and 5.3 per cent (n=3) were boarder/lodgers. Two persons (3.5 per cent) had “other” or ill-defined relationships to the household head. The same general pattern is observed in Type 4 (nuclear family with additional persons) physician households, where, once again, the majority of additional persons (66.0 per cent, n=138) were domestic workers; of those remaining, 22.0 per cent (n=46) were relatives, 9.1 per cent (n=19) were boarder/lodgers and 2.9 per cent (n=6) were “other” persons. While domestics continue to be important contributors to the Type 6 (single-parents and children with additional persons) physician households (35.3 per cent, n=12), some 32.4 per cent (n=11) were boarder/lodgers, 20.6 per cent (n=7) were relatives, and 11.8 per cent (n=4) were “other” persons. The ubiquity of live-in domestics suggests a high level of wealth among these physicians, especially since a number of households had more than one. In the census, most domestics working in physician households reported earnings of between $100 and $150 per year.

The Peters–Primrose duplex on College Street provides two typical examples of the Type 4 household. According to the 1901 census, the Primrose household consisted of 39- and 37-year-old Dr and Mrs Primrose,⁴¹ their three children (ages ten, eight, and six) and three Canadian-born women servants (ages thirty, twenty-three, and twenty-one; who earned between $120 and $144 a year). Their immediate neighbours, the Peters family, included the parents, a baby girl, and two domestics. While the employees are vaguely described in the census as “employed as a servant”, it is possible that their duties—like the responsibilities of some doctors’ wives—extended beyond typical domestic responsibilities to assisting with the upkeep and functioning of the medical office.⁴² In the sources examined for this study, it is impossible to know precisely the involvement of servants in the practice of medicine. Perhaps household financial records, servants’ diaries, or physicians’ autobiographies will shed light on this question in the future.

The increasingly scientific approach among physicians meant that their offices needed to accommodate even more specialized medical equipment, apparatus that might not have fitted or suited the domestic context.⁴⁶ Shumsky, Bohland, and Knox cite large pieces of equipment, such as “sterilizers, X-ray units, diathermy machines or refracting outfits”; collections of smaller equipment which would need to be stored and on-hand, such as “laboratory glassware, microscopes, centrifuges, vibrators, baumanometers, sphygmomanometers, syringes, needles, cotton, gauze, bandages, plaster and gloves”; as well as myriad office and examination room furniture.⁴⁷ Despite the move of physicians’ offices out of the family home, doctors continued to surround themselves in domestic imagery, which retained an important place even in this new professional context. Architects accommodated new medical technologies in medical high-rises, but softened its impact through home-like touches. The Medical Arts Building in Toronto, for example, included a twelve-bed hospital which was made to look as homey as possible: “Patients themselves enjoy the atmosphere of the hospital, which as we have said before, has been made as uninstitutional as possible and yet as scientific as illness demands”, claimed a journalist writing for Canadian Hospital.⁴⁸ It is precisely the tension between domesticity and science which Shumsky, Bohland, and Knox argue transformed the practice venue since “once doctors had perceived ‘scientific skill’ as a ‘commodity’ to be bought and sold, the home-office conjunction could no long continue to exist”.⁴⁹

Some of the same successful doctors who commissioned house-offices were key players in the development of clinical high-rises. In Toronto, 200 doctors participated in the venture to establish the Medical Arts Building; Dr Primrose was the president of the development. Each physician in the organization was part owner of the building and had office space allocated therein.⁵⁰ By 1919 Primrose (soon to be the new dean of the Faculty of Medicine at the University of Toronto) had left the downtown duplex he had inhabited next to Peters, moving his home to 50 Forest Hill Road, near the intersection of St Clair Avenue and Avenue Road, decidedly north of the central core where he had once lived. Forest Hill would become one of Toronto's most prestigious neighbourhoods. This spatial dynamic of downtown office and outlying residence was a new phenomenon observed by Shumsky, Bohland, and Knox for physicians in San Francisco during the transitions taking place in medical practice between 1881 and 1941.⁵¹ And just when the doctor's home left downtown for good, other non-medical typological hybrids appeared too. Modern urban office buildings in the 1920s began to include medical offices. Barott and Blackader's Bell Telephone Company, Beaver Hall Building, Montreal, of 1928–29 (Figure 18), for example, boasted a whole floor of medical facilities.

The move from the intimate house-office to the relatively anonymous, but still exclusive, office tower is an important subject for future research. Shumsky, Bohland, and Knox noted that, in San Francisco, “the majority [of physicians] had ceased to work at home by 1911, and the separation of home and office was nearly universal by 1941”.⁵² The factors they cite to account for this move include: the increase in scale of physicians’ practices to include receptionists, nurses, secretaries, and book-keepers; the need to strengthen links between general practitioners and specialists for the most efficient medical care of patients; the rise in physicians’ income (which made the split possible); improvements to transportation and communication infrastructure; and the change in doctors’ own perceptions of their practices, since many “came to see themselves more as scientists than as family friends, [and] it became more appropriate to treat patients in laboratory-like rather than home-like settings”.⁵³

Physicians drew their insights for group-based medical practice from several models. In Rochester, Minnesota, the Mayo Clinic, a family-based collaboration which began in the 1880s, grew into a dynamic institution with seventeen doctors, eleven clinical assistants, and its own building in 1914.⁵⁴ The Mayo Clinic became a primary model for cooperative practice among physicians, attracting national and international attention.⁵⁵ Cooperative practice was also embodied in the staffing of general hospitals. Hospital employment, however, was a trade-off to the independence which physicians enjoyed in their own practice. It is important to re-state that the doctors involved in the Medical Arts Building bought into the cooperative practice which emerged. According to Madison, hospital-based employment might have attracted specialists because of the presence of expensive but necessary medical equipment, but, by banding together, enterprising independent physicians could secure the same equipment. Further, Madison argues, “for a more technologically intensive style of practice to spread beyond the hospital into the private office, the physician had to adjust his outlook and assume the unfamiliar role of proprietor of a commercial enterprise”.⁵⁶ The timing of the Medical Arts Building may also be tied, to some extent, to another example of group practice outlined by Madison, namely, the “highly organized medical work in the base hospitals” experienced among physicians working overseas during the First World War.⁵⁷ Was it merely coincidental that the president of the Medical Arts Building organizing committee, Dr Primrose, had joined the military effort, lending his surgical expertise to base hospitals in Europe between 1915 and 1917?

Whatever the reason that physicians sought out architects, specialist doctors, business/academic leaders and a handful of architects formed an elite group. In this wealthy and well-educated circle, word-of-mouth could go far in furthering an architect's career. D B Dick, who designed home-offices for Machell (1886), Thorburn (1891), and Peters–Primrose (1898), was commissioned for a number of projects at the University of Toronto, including a post-fire restoration of University College in 1890.⁵⁸ The network of Toronto architects and doctors also played out beyond the city's limits. Burke & Horwood designed for the idyllic cottage country of Muskoka, roughly 160 kilometres north of Toronto. Torontonians fled here for recreation, but also to use the area's tuberculosis sanatoria. Muskoka, like Saranac Lake in New York State, came to represent health and vigour.⁵⁹

The relationships between physicians and architects in turn-of-the-century Toronto extended beyond the design of home-offices. The Primrose cottage (Figure 19) was located in Muskoka and the design included no space whatsoever for his medical practice, suggesting that Dr Primrose chose Burke & Horwood for reasons other than their health care expertise. The same architects also designed a Toronto home for William Gage, publisher, philanthropist, and founding member of the National Sanitarium Association, as well as a number of the Muskoka and Toronto sanatorium buildings. Burke and his associates struck complex connections with the medical community, the university, and philanthropists who supported health care and education.