In 1944, Bruno Mathsson, one of the most important modern furniture designers, disclosed that he had a “great old interest in homebuilding” and joked that he might “change my profession to homebuilder instead of the furniture idiot [möbelidioten].”1 He was thirty-seven years old at the time, and his bentwood chairs with basket weave webbing had been exhibited widely; three models were in the collection of the Museum of Modern Art in New York. He oversaw his workshop busily producing these chairs in his hometown of Värnamo, Sweden. Yet it was no joke—Mathsson indeed changed his focus, and over the next twelve years he introduced few furniture designs but completed about thirty-five buildings in which he introduced several experimental methods of environmental control. His architecture, which has much to contribute to the broader story of mid-century modernism, is little-known outside of Scandinavia, and English-language sources are few.

It is said, more or less correctly, that Mathsson brought the glashus (glass house) to Sweden after visiting Philip Johnson’s glass house in New Canaan in 1949. The following year he built his own furniture showroom, fully glazed on three sides, bringing a new world of transparency to the Nordic country (fig. 1). Within a few years he had mastered his own brand of glass architecture, a significant contribution to mid-century modernism. And he harbored a larger ambition. While Mathsson acknowledged Sweden’s cold temperatures and limited sun, he soon imagined a new utopia of glass: “I believe that the Swedish homes of the future will be a kind of greenhouse with tropical heat and a swimming pool, as pure opposition to our bad climate [dåliga klimat].”2

It is not said, curiously, that Mathsson brought the solhus (solar house) to Sweden, even though he was also strongly influenced by George Fred Keck and the solar house movement of the 1940s. In the major biography of Mathsson (in Swedish), Ingrid Böhn-Jullander revealed that the designer wrote of solar heat (solvärmen) as early as 1944, in the “furniture idiot” passage. This was before he had built a single structure. At this early date, by studying Keck’s work, Mathsson had a thorough understanding of how American insulated windows could admit solar heat during the day and retain it at night. He eventually manufactured his own windows to do just that. Furthermore, while he visited New Canaan, the designer traveled across the U.S. and most importantly to Chicago, where he probably spent considerable time with Keck and his work, as Swedish scholar Karin Winter has written. Yet he never followed Keck’s solar heating methods faithfully, and he never used the word solhus to describe one of his homes.

*          *          *

In Mathsson’s earliest initiative in architecture, he created new methods of wall construction in order to solve what he called “the window problem.” In 1945 he wrote:

Habitual thinking says that a window should be openable, but developments already show that in the near future we will have the hermetically sealed house and room cubes, where ventilation and heat supply take place through so-called “air conditioning.” One may therefore already now ask oneself, whether the openable window is as necessary as we would like to believe?3

Here Mathsson followed a trend in America to dissect a window’s functions (discussed below), which ultimately stemmed from Le Corbusier’s 1929 statement: “la fênetre est faite pour éclairer, non pour ventiler” (“the window is for light, not for ventilation”).

Mathsson possessed a strong technical orientation and his interest in fixed glass was rooted in an engineer’s mindset about environmental control:

When we build a house, what is our ultimate objective? It is really to fixate a certain quantity of air, to separate it from the great atmosphere, and, by our willpower, to regulate the temperature fluctuations ourselves … If house building could rid itself of all stylistic concepts and routine thinking and instead study the actual problems that exist and decide to separate a certain quantity of air in the most efficient way possible, before long we would arrive at the ideal house.4

Mathsson was also “impatient with Sweden’s wooden-house industry” (PGH 124), which, since the 1920s, delivered prefabricated homes using flat-packed elements called plankwall (plankstomme). Mathsson created a “post or skeleton system” (stolp- eller skelettsystemet), similar in concept to American frame construction, and applied for a patent in 1945. The patent application was simply entitled Byggnadskonstruktion (Building Construction).5 (Here he participated in a broader initiative in Sweden—a story yet to be told—to study American construction and modernize Swedish homebuilding by using wood more efficiently.) Mathsson’s system was to be factory-built as panels, and he created details where the fixed panes of glass would fit between the structural posts without a separate window frame. He may have seen this detail in some earlier projects by California architects like Gregory Ain’s Dunsmuir Flats. In any case, Mathsson’s patented wall system would permit a highly repetitive fabrication process, and expressively it would favor a lean, spare aesthetic of repetition in step with 1940s prefabrication more broadly. Ultimately, though, the point of the new framing technology was to solve the window problem by introducing fixed glass.

Mathsson employed his framing system with fixed glass in four significant projects built in Värnamo in 1947–48: two garment factories, a confectionary, and 21 units of workers’ housing in a garden setting. The workers’ housing, for machinists, appears to be particularly notable in retrospect, although it was not well-publicized in its day and remains obscure. A Swedish newspaper described the workers’ housing as “epoch-making for the building market” because fixed glass had not been used in this manner in residences before (MGM 149). The repetitive nature of Mathsson’s framing system with fixed glass bears a clear relationship to 1940s prefabricated housing in America (shown in figs. 2 and 3). His next client studied the Värnamo workers’ houses and found the fixed windows to be “most revolutionary … as you avoid all window carpentry” (MGM 152).

The novelty of fixed windows in Sweden is apparent in the variety of labels used at the time. Mathsson called them fasta glasrutor (“fixed glass panes”), elsewhere they were described as fasta glasväggar (“fixed glass walls”), fasta fönster (“fixed windows”), stängda fönster (“closed windows”), and one newspaper identified them as igenspikade fönster (“nailed-shut windows”). More recently they have also been called icke öppningsbara fönster (“non-opening windows”).

From his earliest architectural works in 1947–48, Mathsson also understood that fixed windows required alternative methods of natural ventilation. It would seem that opening the house contradicted his idea of separating the interior “from the great atmosphere,” but he had probably been persuaded by the emphasis on natural ventilation in the work of Keck and other architects in the United States.6 (Though he never had a formal education in architecture, Mathsson was “avid” in his self-education and subscribed to many European and American architecture journals.) Keck and others used “ventilating louvers” in concert with fixed glass (seen in fig. 4), and this new technology was well-published in the 1940s.7 Mathsson, naturally, did not copy the American models but designed and fabricated his own variants. He installed fixed glass in his furniture workshop and created a variety of “adjustable fresh air intakes” (reglerbara friskluftsintag).

Mathsson grasped another key technological concept from his studies of American solar architecture: the efficacy of solar heating depended on the new technology of insulated glass to keep the heat in. Insulated glass units (multiple panes with sealed gas-filled cavities between the glass) were difficult to source in Scandinavia. Yet Mathsson could sense the revolutionary potential to come. “In America,” he wrote, “the so-called window problem has been solved by the new thermoglass windows, which we do not manufacture here in Sweden. … I have a feeling that an enormous development will take place precisely on this problem and that will completely revolutionize our view of the home, its architecture and interior design” (MGM 136).

*          *          *

In late 1948 Mathsson and his wife Karin left Sweden for a six-month-long visit to the United States at the encouragement of Edgar Kaufmann Jr., director of industrial design at the Museum of Modern Art. (He had collected Mathsson chairs for the Museum in the 1930s and personally purchased one for his family’s home, Fallingwater.) Mathsson was not a tourist; he came to study architecture and new methods of construction. Kaufmann helped plan their itinerary. A few examples indicate the scope and significance of the tour: the couple spent time with Eliel Saarinen at Cranbrook, Walter Gropius and Marcel Breuer at Harvard, Frank Lloyd Wright at Taliesin West, and Ray and Charles Eames, whose house in Santa Monica was under construction. At all his stops, Mathsson deepened his understanding of American construction methods with special attention to the use of glass, ventilation, and methods of floor heating.

In Chicago, Mathsson likely met with Fred Keck, whom I have characterized as “the first solar architect.” By 1949 Keck, assisted by his brother William, had convincingly answered the technical and aesthetic problems associated with passive solar heating and had stimulated the widespread solar house movement of the 1940s (TSH 13–ff, 73–ff). (I also emphasized that Keck and a few other science-oriented solar architects were among a small minority; most architects relied upon mechanical heating and cooling with little regard for energy use.) Given Mathsson’s agenda—to learn about solar heating with fixed insulated glass, ventilation panels, floor heating, and prefabricated framing—visiting the Keck brothers would have been Mathsson’s top priority. In fact, even before traveling to the U.S., Winter discovered, Mathsson had written a brochure “describing the advantages of living in a glass house,” which turns out to be a Swedish translation of a Reader’s Digest article describing Keck houses (PGH 119).8

While in Illinois, Mathsson would have been most interested in the Kecks’ project for Green’s Ready-Built Homes in Rockford.9 The Green’s Ready-Built Homes (see fig. 4) included panelized construction with Thermopane and ventilating louvers, as well as a hollow-tile floor system heated by air as well as passive solar. Over a hundred homes were built in 1945–46. Using scientific calculations, the brothers estimated these features would reduce fuel bills by one-third. It was also one of their most powerful designs, a fully resolved aesthetic statement of environmentally-responsive architecture. Back in Sweden in mid-1949, Mathsson continued to study the Kecks’ work in great depth.10 Clearly Mathsson possessed a full understanding of the scientific principles and aesthetic possibilities of the solar house and was well-prepared to participate in the international solar architecture movement at mid-century.

But the American house that seems to have affected Mathsson the most was Philip Johnson’s glass house in New Canaan (see fig. 5). In an interview with Dag Widman of Form magazine, he held up a picture of the Johnson house and exclaimed: “We could live like this in Sweden too … we are hopelessly backward here in Sweden!”11 He simply “could not hide from delight over living in a glass house.”12

Mathsson must have recognized the essential contradictions between Fred and William Keck’s houses and Johnson’s house in New Canaan. As I have shown elsewhere, “the solar house” and “the glass house” were fundamentally incongruent in concept because the solar house required design priorities that the glass house did not: differential orientation, shading based on solar geometry, and ventilation (TSH 35–ff). Mathsson seems to have visited Johnson’s house in January, presumably during the daytime, and one imagines he experienced high solar gains on a cold day. Did he realize how poorly the house performed—how hard the heated floor and ceiling worked—at night? While the Kecks’ solar houses could claim significant benefits in terms of comfort and energy use, Johnson’s house is now well-known for its poor environmental performance.13

After returning to Sweden, Mathsson seems to have been able to reconcile these contradictory influences with little anxiety. “He pointed out on many occasions how inspiring this journey had been. After that, he became more confident in his choice of entire glass walls and underfloor heating” (MGM 161). Like Jefferson after Paris or Wright after Japan, Mathsson returned a more confident and mature architectural designer, brimming with creative energy. He was “eager to distinguish himself as a modern architect” (PGH 124).

*          *          *

To build glass houses in Sweden, Mathsson needed to create and manufacture new building materials. As mentioned above, he understood the importance of insulated glass units, which were available in America but uncommon in Scandinavia. So, in about 1950 he developed a triple-glazed window insulated with nitrogen. This was a relatively sophisticated design and engineering challenge, but again Mathsson leaned on his experience in fabrication. One of the key members in his workshop later recalled the high level of creative problem-solving involved in this effort.14 While many modern architects created bespoke construction details or participated in factory production, few worked as an independent manufacturer with this level of technical depth. He called the product “Brunopane.”15 The units were considerably thicker than the triple-pane glass produced today; Brunopane occupied the full depth of the structure, with a double-pane section outside and the third pane inside (see fig. 6). When he needed to explain the science of Brunopane to the public, Mathsson said that insulated glass worked in the same way as a thermos.

In his first major project after visiting the U.S., Mathsson built the first structure in Sweden that could rightfully be called a glass house (glashus): his own furniture showroom (Värnamo, 1950, see figs. 1 and 7). The room featured Brunopane triple-glazed panels on three sides of the main space, and for the first time the glass stretched from ceiling to just above the floor. Whereas Mathsson’s previous projects still conveyed the character of a wall with windows, this project represented something new—the thrilling transparency of the glass house. The originality of this space alone is enough to secure Mathsson’s place in the history of Swedish architecture.

Despite the place New Canaan held in his memory, Mathsson departed from Philip Johnson’s example in that he never built a four-sided glass house nor a glass wall without an overhanging roof. And when he began to build with Brunopane he did not mimic Johnson’s use of undivided plate glass. He created smaller Brunopane units and built them into a wood framing system to create a glass wall divided horizontally and vertically. Mathsson’s typical wall had a three-part composition, a large viewing pane in the center, horizontal panes above and below, like Keck’s houses in Rockford. To panelize the glass walls for factory production Mathsson used a one-meter module, uncommon in Swedish homebuilding, which he said was inspired by the Japanese tatami mats (FMN 20). The idea of a Japanese-inspired divided glass wall was likely derived from his familiarity with California modern houses and validated by his visit to the Eames house.

Yet in designing the showroom Mathsson did not strictly follow Fred Keck’s solar house formula, which prescribed south-facing glass and opaque walls in the other three orientations. While the Mathsson furniture showroom had a long glass wall properly facing south (see right side of fig. 1), it had the identical glass wall on the north, which would be absolutely wrong in Keck’s eyes because this design, even with Brunopane, would produce far more heat losses than heat gains. In the end the scheme borrowed from both Johnson, Keck, and other sources, demonstrating Mathsson’s ability to assimilate a complex variety of influences to arrive at new forms and new aesthetic effects.

The furniture showroom was built on a “concrete slab directly on the ground” (betongplatta direkt på marken) (BIG 29),  as Mathsson had observed in America. (The predominant method in Sweden was a wood-framed floor raised above a crawlspace or basement.) Then, the design included a hot-water system with pipes cast into the concrete slab, possibly the first of its type in Sweden. This meant radiators, the traditional norm in Swedish homes, could be eliminated and the glass could come down to the floor. Though detailed information is limited, it appears that Mathsson designed a novel type of piping system, inspired by American heated floors but inventive in some manner, and later pursued a patent.16 No insulation was provided under the slab at this time, but the heat loss into the soil was seen as a positive feature because “you can build directly on the ground” without building foundation walls to the depth of the frost line (BIG 29–30). Again, new vocabulary was required; Mathsson’s slab-on-grade structures became known as “basement-free houses” (källarlösa hus) (BIG 32).17

Mathsson realized that this type of floor construction offered a new experience of space, a seamless relationship between indoors and outdoors. One inhabits the landscape rather than observing it from above. He spoke of “coming down to earth” (PGH 128–35). Inside the room, the innovative environmental features disappear, as does the structure of round steel columns painted white. The low-sloped ceiling finished in bare spruce gives the space a domestic character while the marble-tile floor is neutral and durable, another departure from Keck or Johnson influences. For all its experimental vigor, the room conveys the relaxed confidence of a designer and craftsman whose ideas had fully matured. By 1951 he spoke from a new position of authority: “The technical problems are largely solved.” Therefore, he issued a social challenge: “Now it is time to convert the Swedish people, including authorities and trade unions, to this new way of building and living, so that those of us now living, not just our descendants, will be able to enjoy tomorrow’s home” (BIG 29).

*          *          *

Mathsson built a similar showroom for the Kosta glassworks in 1953–54, and then for the company’s workers he created a five-unit housing project (now called Kosta Glashus), recognized then and now as a major work of Swedish architecture. Each unit was identical and L-shaped in plan to define a private outdoor space for each unit which he called a solgård (“sun court”). While Swedes certainly cherished gardening in the sun for many generations, Mathsson here seems to have used the word solgård in a new way: to denote an architecturally-defined space for sunbathing and with visual privacy for nudism. If so, the solgård should be understood as a new architectural form in Swedish domestic architecture.

For the Kosta Glashus, each L-shaped unit plan had one leg occupied by bedrooms, with glass facing south onto the solgård. The other leg, for living, dining, and kitchen, had glass walls facing both east and west. This orientation was “incorrect” according to Keck’s solar house principles, a conundrum that will be discussed further below. In total the walls in the workers’ houses measured only about 40% glass, similar to the Kecks’ homes and obviously far more enclosed than a glass house. Indeed, from many perspectives, one would describe the Kosta Glashus as a brick building. In each unit, the north wall of the bedroom wing was fully opaque so that residents would not look onto their neighbor’s solgård.

A 1951 Swedish feature on Mathsson asked: “Doesn’t a glass house become a hotbox in the summer?” And the answer: “No, not if the roof has a sharp projection [kraftigt sprang] and prevents the sun from lying on [the floor]” (BIG 32). Again, Mathsson’s deep familiarity with the solar houses of Keck and Keck (and others) meant that he recognized the environmental importance of proper shading, which Philip Johnson and other high-modernists refused. Figure 8 shows the use of deep overhanging eaves (also called takutsprånget or takfötterna). He explained to Swedish journalists that the roof overhangs would protect the homes from too much sun, just as Keck had emphasized in American publications several years earlier.

Mathsson also improved the amount of insulation in his structures through the early 1950s. This was neither original nor visionary; it reflected the diffusion of new knowledge in the building industry and the availability of new products. For the Kosta Glashus, the floor was again heated with hot water pipes, with all five units sharing a central boiler. But between the heated floor and the structural slab, Mathsson now inserted five centimeters of wood-wool insulation board (träullsplatta). In the roof, a 2006–07 renovation revealed twenty centimeters of mineral wool, much better-insulated than expected (KG 70, 81). Also during that project, some of the original Brunopane units were preserved for their historical importance, and the building’s history was thoroughly documented. After the restoration, the Kosta Glashus was designated a legally protected building monument.

Through a series of houses for private clients in the mid-1950s, Mathsson introduced no new innovations but deployed his now-established menu of environmental features in a fascinating variety of site-specific responses. Each included a heated concrete floor at ground level, generous areas of triple-glass Brunopane walls, and separate devices for natural ventilation. The solgård plan—L-shaped, where glass walls facing solgård are oriented to the southeast and southwest—became Mathsson’s favorite configuration. Most rooms, including bedrooms, were given glass doors leading directly outside, underscoring the priority of indoor-outdoor movement. The 1957 house for jazz singer Alice Babs in Saltsjöbaden (just outside Stockholm) even included a swimming pool in the solgård.

Mathsson also built four preschools in 1952–54, all similar in character and construction to his residential projects. These were described as “American-influenced” in a recent historical survey,18 and as a group they beg for future research and interpretation. (And like most his houses, the schools seem not to have been professionally photographed.) Continuing his commitment to factory production, Mathsson also built a house in 1957 which was fabricated in the Värnamo workshop and sent by truck to Rotterdam in the Netherlands. Builders continued to find Mathsson’s designs “surprisingly cheap” to build (KG 33–34).

During this period Mathsson shifted his method of heating floors from hot water to electricity; he said it was “an even better and simpler solution” (BIG 29). He found a Norwegian engineer, Georg Jacobsen, with experience in electric heating of greenhouses to assist him with calculations and wiring design. This prompted a wider discussion of domestic energy issues in Sweden. Electricity was an attractive source of energy in the early 1950s due to high fuel prices. (Sweden’s energy use was dominated by fossil fuels in this period; only about 10% came from hydropower and nuclear was introduced later, in the 1970s.)19 In two different period publications it was acknowledged that widespread adoption of electric residential heating would create “a problem in the long run” for the country’s capacity to generate electricity (BIG).20 Apart from the question of electricity vs. hot water, Mathsson’s insight is fully validated in Swedish homebuilding today; the slab-on-grade (platta på mark) construction method with radiant heating (and robust underslab insulation) is practically standard.21

*          *          *

From a scientific perspective, was it irrational to bring the Johnson-style glass house to Sweden? Of course. It was a poor choice in New Canaan and would be worse in a colder and more northerly location. But was it irrational to bring the Keck-style solar house to Sweden? This is a more complicated question. Värnamo is situated in the Småland region of southern Sweden, closer to Copenhagen than Stockholm. It is both significantly colder than Chicago and has a lower solar resource, as shown in the tables below.

Värnamo is colder than most locations in the lower U.S., comparable to Fargo in terms of heating degree days. (Heating degree days are =a measure of how much heating is needed for indoor comfort. A higher number means a colder climate; more heating is needed.) Due to the northerly latitude, all of Sweden has a poor solar resource in the winter months. Värnamo’s latitude is 57°11’N, about the same as Juneau, Alaska, and further north than all major Canadian cities. To put a finer point on it, at the winter solstice in Värnamo, sunrise is at 8:34 am and sunset is at 3:10 pm, plus it is likely to be cloudy. While the solar picture improves in March and April, when significant heating is still needed, it is still worse than Chicago. Considering the data above, passive solar heating is a difficult sell in Sweden. It is understandable that Mathsson did not promise his clients significant energy savings due to solar heat, as Keck did.

Many other Mathsson projects, besides the furniture showroom mentioned above, were oriented incorrectly according to Fred Keck’s solar house principles and scientific thought in the context of North America. Similarly, the principal glass façade at Villa Prenker (see fig. 10) faces north-northwest and at Villa Södrakull it faces west (see fig. 12). In each of these cases. a well-insulated opaque wall would perform much better, particularly in such a cold climate, but Mathsson had other priorities. (A few of his houses used a Keck-like linear plan with all rooms facing south.) What about Mathsson’s favorite orientation, the L-shaped plan with glass walls facing southeast and southwest? This would be “wrong” in the U.S., but in Sweden the question is somewhat more complicated.

The data above, showing that the most beneficial solar heating occurs in the swing seasons (spring and fall) but not winter, may also help explain why Mathsson did not draw cross-sections showing solar angles and shading. Such diagrams normally show the extreme geometric conditions: the highest solar altitude, at noon on the summer solstice when the window is to be shaded, and at noon on the winter solstice when the window should be in full sun. But what if you don’t expect much sun on the winter solstice and you don’t mind some solar heat gain in the summer? Which angles would you depict if the ideal heating condition were mid-morning in March, or late afternoon in September? And for a building facing southeast and southwest, rather than south, it is almost impossible to imagine a simple diagram that would make sense.

Does all of this mean that Mathsson did not seek to optimize his architecture for heat gains and losses and to save energy? In his pursuit of solar heating, it was “outside his ambitions to study technological science in a stricter sense,” according to Elias Cornell. So, he consulted with Gunnar Pleijel, a research faculty at the Royal Institute of Technology in Stockholm, and “the best theorist for sun and heat among our architects.” Cornell recalled: “In their discussions, the practitioner’s intuition and the theorist’s assumptions clashed.”22 While the exact nature of these clashes is not clear, the word “intuition” emphasizes a key point: Mathsson did not understand the sun and the climate principally through scientific data (like the numbers discussed above and below) but through his senses and his experience.

Given Sweden’s lack of sunlight in the winter, and given the fact that there are significant heating needs in the swing seasons, is it possible that a south-facing L-shaped plan with glass walls facing southeast and southwest, Mathsson’s favorite, is in fact as sensible as the Keck-type linear plan with south-facing glass? Indeed, the answer is yes—almost. In 1963 Pleijel published data (for Stockholm but likely quite similar for Värnamo), which showed positive or negative balances (heat gains compared to heat losses) for eight window orientations over the twelve-month calendar. In the four months from November through February, all window orientations are negative due to low solar gains and cold outdoor air. But in the other eight months (totaled), southwest-facing windows are 99% as beneficial as south, and for southeast-facing windows the figure is 90%.23 Thus little penalty was paid for departing from the American rules about solar architecture, and he likely did more good in the mornings and afternoons in the swing seasons. The risk of overheating would be rare. Perhaps his “intuition” was better aligned with the science than we might assume.

Still, Mathsson’s inconsistency about orientation, particularly his use of north-facing glass walls, seems to contradict some of his stated ambitions. Although Mathsson offered incisive quotations throughout his architectural career (many included above and below), in retrospect his philosophy about the interior environment and energy use turns out to be fairly slippery. This is due, in part, to the fact that these quotations come from scattered interviews and other bits. To read and analyze his work would be more straightforward if he had written a polemical or reflective piece about his buildings and ideas, but he never did.

*          *          *

When we speak of architecture and “the environment,” there are two distinct traditions inherited from the nineteenth and twentieth centuries. The first is principally scientific, concerned with controlling the environmental conditions through passive and active strategies. This includes those enthusiastic for “applied power,” like Reyner Banham in The Architecture of the Well-tempered Environment, and those who advocated for energy-efficient methods like Fred Keck. The second is principally therapeutic and psychological, concerned with sunlight and fresh air, transparency, and seamless indoor-outdoor movement. This tradition was incubated in the Arts & Crafts movement (with a heavy dose of Japanese influence), confirmed in early-twentieth century sanatorium architecture, and fully-formed in the residential works of Schindler, Neutra, and their followers. Both traditions became constituent facts for mid-century modern architects.

Clearly, Mathsson embraced both traditions, but he identified more strongly with the second. Despite his “engineer’s mindset” as characterized above, he spent little effort quantifying heat gain or energy savings. His interest in the environment was more psychological than scientific. According to Arthur Hald, Mathsson spoke of the psychological effects of transparency and creating an “intimate connection to the terrain.” He contemplated: how would it feel to live in a glass house? Mathsson spoke of “excellent contact between indoor and outdoor spaces [utomordentlig kontakt mellan inne- och uterum]” (BIG 29–32). Of course the central importance of indoor-outdoor transparency to mid-century architecture is well-understood. “Glass has psychological as well as practical value … it can bring the whole outdoors right into the house,” according to Elizabeth Mock in Tomorrow’s Small House.24 Other popular publications, like Tomorrow’s House by Nelson and Wright, promoted the same idea. Further, for some mid-century modern architects the relationship of indoors and outdoors was deeply psychological. Richard Neutra, for example, used sensory and spatial effects as a curative for birth trauma and other emotional problems.25

To understand that Mathsson identified more strongly with the second tradition, it is most useful to study his last projects, those he built for himself and Karin: a summer house called Frösakull (near Halmstad, 1960) and Villa Södrakull (near Värnamo, 1965).26 Frösakull is of particular interest because it was barely enclosed and offered a minimum of environmental control (see fig. 11). Located in a small beachside neighborhood on Sweden’s west coast, Frösakull was as different from a traditional Swedish home as can be imagined. The structure consisted of a simple framework of slender steel tubes enclosed by glass or wood walls and a roof of round vaults made of corrugated plastic, translucent and uninsulated. A home for leisure and relaxation, it pushed the modernist idea of transparency between inside and out into a new realm. Helena Mattson has interpreted it as an “anti-house” and compared it to Reyner Banham’s famous environment bubble (HNH 201).

As usual he designed an L-shaped plan, with siding glass wall sections facing the solgård. Since Frösakull would only be used in the warmest months, the floor was not equipped with a heating system. Mathsson said the house would be heated by solar energy: “First and foremost, I want to use the heat and energy from the sun. In this respect I have gone one step further by making the roof out of transparent corrugated and curved plastic” (HNH 191). While this material certainly created a bright and evenly-lit interior by day, whether it permitted much solar heat is questionable. With the southeast-facing rooms being warm in the morning, middays spent in the outdoor solgård or at the beach, and the southwest-facing living room being warm in the evening, it seems that the house embodied a kind of ritual climate-seeking—“you constantly move after the sun” (FMN 35). To underline the theme of mobility-for-comfort, the Mathssons did not build a permanent kitchen. Instead, Bruno designed a stove mounted on wheels so that food could be prepared in an informal manner as needed.

In the context of the history of Swedish architecture, Frösakull is practically sui generis. With his U.S. trip eleven years in the past, the influences of Keck and Johnson were barely perceptible. And it appears to have had little subsequent influence, having been little-known until recently. According to Karin Winter’s analysis, “The Frösakull retreat is fascinating and poetic, perhaps it is more an intelligent composition of materials than an actual house. Its great quality is not as an experimental house of the future but as a structure supporting perfectly decent idleness and non-activity” (PGH 175).

Frösakull can also be understood as a “health house” meant to support Mathsson’s free-thinking ideas about fitness and outdoor living. He practiced and advocated for sleeping outdoors, nude sunbathing (naturism), and a vegetarian diet. Of course, the healthy influence of sunlight was a well-established scientific fact. On sleeping outdoors, one imagines that the couple experienced some cold summer nights at Frösakull. In general, Bruno slept outdoors between March and October, in a special bed of his own design, with a zippered enclosure and a five-liter hot water reservoir (HNH 198). According to Winter, Mathsson gave radio interviews where he spoke not only of architecture and technology “but also of a new lifestyle, of more liberated, happier people” (PGH 140). In bringing these concerns to architecture the Swede followed a lineage of practice closely associated with California architecture, from Charles Keeler to Irving Gill to Philip Lovell and his architects, R.M. Schindler and Richard Neutra.

More to the point, Frösakull shared a remarkable number of distinctive characteristics with the earlier Schindler house (Los Angeles, 1922). The canonical Schindler house included slab-on-grade construction offering easy indoor-outdoor movement, L-shaped plans defining outdoor rooms, Japanese-inspired divided glass walls including sliding sections, and outdoor sleeping—all features Mathsson adopted. Both Frösakull and the Schindler house related more closely to a camping shelter than a permanent edifice; each architect spoke to this. (Further, Mathsson grasped Schindler’s idea that a camper’s “open front” required a “protected back.”) And in both, the architect used the notion that kitchen work was burdensome to create new form. It is unknown whether Mathsson visited the Schindler house while in Los Angeles in 1949, and Schindler is never mentioned by the handful of Swedish writers who knew Bruno well. Still, it is apparent in retrospect that Frösakull, and the solgård form generally, is more strongly connected to California than Chicago or New Canaan.

Because of its sense of impermanence, Frösakull also speaks to the expanded field of architecture in the modern period. The house’s flimsy materiality and spartan services were practically antithetical to the traditional understanding of architecture as the creation of decorative and permanent monumental structures for elites. To be sure, Frösakull was a leisure home and therefore a statement of the Mathssons’ privileged economic status. But as an “anti-house,” it also reflects these specific social and indeed political commitments: the reorientation of the discipline to common people and their activities. Frösakull’s impermanence was highlighted after Bruno Mathsson’s death, as it was abandoned and left to decay, and poignantly captured by photographer Mikael Olsson.27

The Mathssons’ permanent home, Villa Södrakull, would be his last project in Sweden and a true masterwork of mid-century modernism. By this time, he had essentially retired from architecture and returned to furniture design, largely due to frustrations with local building officials who did not understand his building methods. Here the architect discarded his favored L-shaped plan and used a simple rectangular form due to the nature of the site. The living room and the principal glass wall faces west (see fig. 12), to a broad view of Lake Vidöstern. Still, true to his commitments, Mathsson created a solgård on the south, enclosed by a garden wall.

At Villa Södrakull Mathsson amplified the indoor-outdoor relationship through a highly original use of materials: a glass mosaic floor in emerald green. He had introduced this treatment earlier at the Kosta Glashus (using Kosta glass), where it connected the indoor environment with the grass of the solgård. And it proved to be perfectly compatible with a heated floor. Because Södrakull included a translucent roof of corrugated plastic, over the kitchen, its green floor provokes an even more powerful psychological effect, that of being in a greenhouse. (Recall Mathsson’s desire that “Swedish homes of the future will be a kind of greenhouse.”) The Josef Frank-designed curtains complete the effect (see fig. 13). According to longtime friend Dag Widman: “There was a peculiar calm about this house … and nature seemed palpably close.”28

*          *          *

Although Mathsson’s architecture remains little-known outside Scandinavia, it bears renewed attention today because it is increasingly important to scrutinize historical architecture through the lens of environmental control and energy use. On the one hand, the effort to understand “the relationship between architecture and climate is … of immense practical value,” according to Dean Hawkes, because some old buildings may embody wisdom about heating, cooling, and lighting applicable to future design practices.29 On the other hand, many twentieth-century buildings were “oblivious to climate” due to enthusiasm for mechanical systems and are therefore poor examples for future design practices.30 For both reasons, we must pursue a richer understanding of historical building performance, including empirical information.

And furthermore today, the ethical status of glass architecture is in serious dispute due to lines of critique motivated by the climate change emergency. This discourse is strongest in the United Kingdom, where researchers and practitioners such as Alan Short, Henrik Schoenefeldt, and Simon Sturgis have argued for strict regulations on building all-glass towers. Wired magazine put it bluntly: “Towering glass structures are an environmental nightmare. And there’s a growing consensus that we should stop building them.”31 It remains true, as noted in The Guardian, that “even triple glazed windows lose far more heat than a well-insulated brick wall.”32 This is as true in Sweden as in the UK, and more consequential. Similarly, for single-family houses, most experts say the best strategy is to focus on superinsulation and airtightness, and to ignore passive solar heating. The Swedish kataloghus industry takes this approach with very good results.33

While it is problematic to retroactively judge historical architecture by new values and standards, it is fair to present new insight, and in this spirit, it is clear that Mathsson’s homes would have used more heating energy than a comparable Swedish home with well-insulated walls and smaller windows. Ignoring the therapeutic and psychological effects, floor-to-ceiling glass does not usually make sense in Sweden, then or now. Nor does it appear realistic that “Swedish homes of the future” would be “a kind of greenhouse with tropical heat” without an ample amount of applied power. Of course, Mathsson did not ignore the therapeutic and psychological effects, and consequently he did not focus on energy efficiency. He never calculated gains and losses for his designs as Fred and William Keck did. Nor did he promise his clients they would save on their heating bills. Despite his interest in solar heating, it is clear that Mathsson regarded mechanical heating as abundant and capable of delivering comfort, an attitude about architecture and energy which he shared with most Western architects at mid-century.