In compiling this atlas of urban interiors, I have benefitted from the information provided in the indexes of atrium buildings produced by Richard Saxon (Atrium Buildings: Design and Development, Architectural Press, 1983 & 1986; The Atrium Comes of Age, Longman, 1993), and Michael Bednar (The New Atrium, McGraw Hill, 1986), as well as the database Birkhauser Building Types Online, and searches on indexes including the Avery Index to Architectural Periodicals and the RIBA Library Index.

While there are many overlaps in entries between this atlas and the indexes of Saxon and Bednar, I took a narrower view of what defines an atrium building. What was interesting to me was gathering a list of buildings where the atrium appeared in a way that was not necessarily central to the development of any specific building type, yet could show a consistency of spatial development across types. To reinforce this line of thinking, and unlike Saxon and Bednar, I excluded shopping malls because they constitute a building type with their own specificity and developmental trajectory. Museums and galleries were also excluded because of the inherent typological need for a ceremonial space for organising large groups of people. The consistency seen in the oeuvre of some architects makes these decisions seem difficult and strange. For example, I.M. Pei’s Grande Pyramide du Louvre, Paris (1989) is excluded, as is his East Building of the National Gallery of Art, Washington (1978), a seminal museum atrium. However, the atlas includes several other Pei buildings, including the Dallas City Hall (1978), John F. Kennedy Presidential Library, Boston (1979), Collins Place, Melbourne (1982), and the Weisner Building, MIT (1985). Some bank chambers were included, such as Estudio Sanchez Elia Peralta Ramos and Clorindo Testa’s Bank of London and South America, Buenos Aires (1966) even though these can sometimes be seen as developments of earlier models. 

There are no doubt going to be glaring omissions and contested inclusions. And some boundary cases might seem problematic. For example, Louis Kahn’s Philips Exeter Academy Library is included, whilst James Stirling’s History Library at Cambridge University is excluded. The latter was considered to be an extension of the idea of the reading room as discrete space, albeit with a structural system and form more reminiscent of other contemporaneous atrium buildings. The former organises the library around an atrium which is well in excess, spatially, of what could be argued as Stirling’s typological development of the reading room. 

The atlas is organised in two visual formats: a timeline and a map. The timeline allows examples to be scanned and situated temporally. The map adds a dynamic, geographically searchable way of understanding the location of atrium buildings, something not possible with earlier published lists. Zooming out of the map enables the global spread of atrium buildings to be seen. Zooming in then enables buildings and their atriums to be situated relative to present-day surroundings.

When compiling the atlas, the names of the buildings and architectural firms were taken from the time of the buildings’ construction. These both may have changed subsequently. Dates are at completion of construction or opening of buildings. While the atlas’ start date of 1960 makes an argument about the emergence of the atrium as a new kind of space in architecture from that time, the end date of 1990 is largely arbitrary and determined by available resources to undertake the compilation of the atlas, though it also relates to the historical interests of the larger research project to which it is related. All latitude and longitude data were searched and entered manually. Some of the buildings listed may no longer be extant, so the map reference is to their former site. Exact locations of some buildings were difficult to determine, so reference is given to the nearest possible location, such as a town or urban centre, or campus.