“And because the use of the photographic device is impossible without a proper card catalog … ”: The Typological-Stylistic Arrangement and the Subject Cross-Reference Index of the KHI's Photothek (1897–1930s)

This article is focused on the historical development of the classification system of the Photothek of the Kunsthistorisches Institut in Florenz (KHI) between the time of its establishment in 1897, with the typological-stylistic arrangement mainly by artists and locations, and the systematization of the subject cross-reference index (card catalog) in the 1920s to1930s. Specifically, these two incisive reorganizations of the Photothek were linked, on the one hand, to relocations and related extensions of space for the growing collection of books and images, and, on the other, to the scholars with their different research interests. The evolution of the categories that determined the thematic ordering of the images and their spatial configuration in open shelves for researchers, as well as the practical use of the typological-stylistic arrangement, shows the Photothek as an eloquent testimony of the constant variations in art history theories and methodologies.


Introduction Origin and sources
The Photothek of the Kunsthistorisches Institut in Florenz (hereinafter KHI or the Institute) underwent several reformations in the period between its organization in 1897 up to the 1930s that were related in particular to the expansion of its collections and the resulting changes in the Institute's spatial configuration. The various stages of development can be revealed today on the basis of several sources: from the annual reports drawn up by the directors that provide detailed information on the Institute and its holdings of books and images, and also from documents, records, accounts books, and letters preserved in the archive of the KHI. These are important sources for reconstructing the collections of literature and images that were assembled over the years as a special library on Italian art. Yet the Photothek itself also serves as a most important reference. From the beginning, it was cataloged according to a classification scheme that determined structured scholarly access to images through various finding aids. The system of classification and the shelving of the holdings were modified several times, though they can be largely reconstructed thanks to the inventory books, the card indexes, and the notes on the cardboard supports of the images themselves. In addition, photographs that were taken of the Institute's premises are highly instructive and particularly valuable as pictorial documents. They serve as supporting evidence and as a supplement to the written records, and they also show the space capacity and the way space was structured, thereby allowing us to infer how the boxes with photographs were arranged and practically handled.
In this article, which aims to examine the evolution of an exemplary image collection of a research institute, I will trace the historical structures of the evolving department of the KHI and consider these in light of its classification system-in the sense of the relationship between photography and photo archive as suggested by Tiziana Serena and, with reference to theoretical approaches to archival research by, among others, Elizabeth Edwards and Geoffrey Batchen: "The photo archive represents the key to understanding the photograph, even in its materiality, and equally needs to be understood in its materiality and social implications … ." 1 To which one could add: and vice versa.

The Founding of the KHI
The opening of the Kunsthistorisches Institut in Florence in the year 1897 was preceded by a long-term period of development that was paralleled by the institutionalization of the still young discipline of art history at the universities. 2 As early as the 1870s, the idea to enable the study of Italian art in close proximity to the original works of art and architecture, at an institute featuring a library and image collection geared especially to this object of study, had taken on concrete form among the "German colony" in Florence. This circle of artists, men of letters, art historians, and collectors-including, among others, Karl Hillebrand (1829-1884), Adolf von Hildebrand (1847-1921), Hans von Marées (1837-1887), Arnold Böcklin , and Adolph Bayersdorfer (1842-1901)-continued the tradition of the circle of figures, such as Carl Friedrich von Rumohr (1785-1843) and Johann David Passavant (1787-1861), that had formed in the 1830s and, starting in the 1860s, had been grouped around the collector and art expert Baron Karl Eduard von Liphart (1808-1891). "For the small colony of German and German-minded travelers who wanted to enjoy and study Italian art at its source in Florence his (Liphart's) inviting home was, for more than a quarter of a century, the hub," as Wilhelm von Bode (1845Bode ( -1929 noted in describing the house on Via Romana and its ambience: "Inside it looked curious, especially in the study. Portfolios full of the most valuable old prints and drawings were lying on and underneath the tables and chairs, interspersed with photographs, plaster impressions and books for which there was no space in the repositories and on the shelves." 3 By that time, August Schmarsow (1853-1936) had already held a seminar in Florence with his students from Breslau (present-day Wrocław in Poland), including Aby Warburg (1866-1929), Hermann Ulmann (1866-1896), and Max J. Friedländer (1867-1958), and proclaimed with this "pioneering feat" in 1888 the establishment of the Kunsthistorisches Institut in Florenz. "That was a purely notional act, for an institute in the material sense was not yet founded at the time." 4 Five years later, in 1893, an official committee was formed that was to encourage the plans, and in 1894, the committee published a "Call for the founding of the art historical institute," asking for donations. After the founding, in November 1897, the honorary director, Heinrich Brockhaus (1858Brockhaus ( -1941, provided a room in his private home at 19 Viale Principessa Margherita (today 43 Viale Spartaco Lavagnini), in which books and images as well as plaster casts were available to scholars ( Figure 1).
The friends' association established in 1898 and with financial support provided by the German Reich from 1902 onward allowed the Institute to expand by renting an adjacent residence and to hire Otto Wenzel as a library assistant. 5 Through Schmarsow and Brockhaus, the KHI was closely connected to the University of Leipzig where, in 1873, one of the first chairs of art history was created-other early chairs were founded in Strasbourg (1872) and Bonn (1873)-and where Brockhaus (in 1892) and Schmarsow (in 1893) were appointed professors. 6 Another connection existed through the fact that the basis of the collection of books and especially images came from Hermann Ulmann, who, as a participant in the 1888 seminar, was among the founders of the Institute. 7 Heinrich Brockhaus described the collection in 1902/1903 in the first annual report of the KHI: "The current holdings of the image collection add up to about 11,600 sheets. This includes the books and images from the estate of Dr. Hermann Ulmann, who died in 1896; those holdings had already for years been loaned to the Institute by the family." 8 Apparently, Ulmann had, at the same time as the images, also provided the Institute with a proper piece of furniture, as suggested by a list of furniture that was drawn up by Brockhaus in preparation of the 1912 move: "a brown picture cabinet from Dr. Ulmann." 9 The foundation of the Florentine image collection thus consisted of the reference library of an individual scholar who had collected literature and pictorial material on his own initiative according to his particular interests and compiled a scholarly apparatus-much like the (later) private collections of, for example, Bernard Berenson (1865-1959), Aby Warburg, and Federico Zeri . Among Ulmann's images are primarily photographs of paintings and drawings-his specialty-followed by about half as many images of sculptures. By contrast, they included fewer than one hundred architectural images, even though large numbers of high-quality photographs of architectural objects were produced, especially in Italy. 10 The quantity ratios thus clearly reflect the interests of Ulmann, whose dissertation on Sandro Botticelli (1445-1510) was published in 1893, as well as the state and nature of scholarship focusing on Italy within his circle.
In the Institute's early years, the holdings of books and images continued to grow through gifts and purchases and photo campaigns made possible by donations; beyond the huge demand for images and books on Italian art of the Middle Ages and the Renaissance in general, there was, at first, no attempt to predefine a more clearly focused collection profile. The collection of reference works being assembled was indeed in need of everything: from standard references and general surveys to specialized literature and images; 11 one could only start thinking about a less passive and more active acquisition policy when a critical mass was reached. Especially in the case of a nascent collection, new accessions tend to be dependent in their makeup on gifts and donations, rather than on actual needs, and chance or coincidence plays a certain role. 12 The donations of books and images, or funds for the acquisition of them, were itemized in the annual reports, and donors were mentioned by name, which allows us to infer a certain correlation between the donors' research interests and the admittedly heterogeneous makeup of the image collection. Already in the annual report of 1904/1905, mention is made of a photo campaign that had been made possible through a donation by Franz I, Prince of Liechtenstein (1853-1938), who had close ties to the KHI. In the Institute's Wissenschaftliche Besprechungen (scholarly discussions), Brockhaus had introduced the "artworks in the environs of Florence near Bagno a Ripoli," which were first mentioned in the 1904/1905 annual report with a note that "the paintings have since been photographed thanks to the generosity of His Highness Prince Franz von Liechtenstein." 13 Brockhaus further described the meetings: In the past years the Institute has repeatedly served as a meeting place to serve the interests of users by organizing small sessions that, rather unambitiously, are called scholarly discussions. … In those sessions rather brief but still very interesting remarks alternated with official lectures depending on the topic and as requested. In many cases these were made even more worthwhile through the showing of photographs, particularly such that were taken by the scholars themselves and not commercially available. 14 Still, the atmosphere was not like at Liphart's home where "those that came early could still take a seat at the round table in the anteroom and have a cup of tea with the older gentlemen; yet the very first question prompted a discussion and books and folders were brought out, the tea had to be cleared away, for the 'Liphart evening' began." 15 It can be assumed, though, that the images were passed around or put on display, for only the 1913/1914 annual report refers to the purchase of a "photographic projector along with a large number of photographs (more than 1,000 items)." 16 That the device was being used is shown by a cash book entry from February 20, 1915, which lists the expenses for the use of the projector as 5 lire. 17 The notice in the annual report also offers evidence that photographs were frequently taken by the members of the Institute themselves and probably found their way into the image collections later on. However, before becoming available in the image collection, they first needed to be mounted on cardboard, as noted in a commentary in the accession inventory of the Ulmann estate: "272.40 lire to Brogi for mounting 1,286 photographs on 1,284 sheets (November 8, 1898)." 18 The "Image Collection" of the KHI In addition to an account of the activities of the Institute, its financial situation, and the activities of the association (with a roster listing the names of the board and members), the annual reports of the KHI also included a survey of the Institute's collections: the library and "image collection." 19 Up to the 1912/1913 annual report, the latter is consistently referred to as Abbildungs-Sammlung (image collection) or Sammlung der Abbildungen (collection of images), 20 because the collection was indeed made up not just of photographs, but also of various kinds of prints.
Brockhaus's successor, Hans von der Gabelentz-Linsingen (1872-1946), 21 did adopt a name for the collection, but in the years that followed the images received scant mention, which may be related both to the fact that the Institute was closed from 1915 until 1923 due to World War I, as well as to a shift of focus areas. The relocation of the Institute to the Palazzo Guadagni at 10 Piazza Santo Spirito in November 1912 offered the opportunity to reorganize the collections. The space capacity and the structure of the space invariably informed the KHI archive and classification system, a fact that should be considered in terms of the "spatial turn" 22 -in a narrow sense-with a view to maintaining the system of open shelves. At that time the holdings included approximately 7,800 books, about 27,600 cardboard-mounted images, as well as 1,662 plaster impressions of medals. 23 The library's new classification system was published in the 1912/1913 annual report. 24 There were obviously plans for a similar rearrangement of the image collection, notably a separation of photographs and prints, as suggested by a remark in the 1925 Festschrift for Wilhelm von Bode. 25 This was probably due to the huge increase in image material and the changes caused by this growth to the systematic classification system, and, finally, to the Institute's move into larger premises, which had become necessary. Among others, the Stuttgart-based scholar and long-time board member of the association, Cornelius von Fabriczy (1839-1910), had bequeathed approximately 3,000 images of works of architecture and sculpture to the KHI that were subsequently inventoried, little by little. 26 Heinrich Bodmer (1885Bodmer ( -1950, who was appointed director in 1922, was the first to refer-in his address marking the reopening of the Institute at the Palazzo Guadagni on October 15, 1927-to a Photographien-Sammlung (photographic collection). 27 Yet the current name, Photothek, does not appear in the annual reports in the period up to World War II. Only after the Institute was once again placed under German administration and officially reopened on October 7, 1953, the newly appointed director, Ulrich Middeldorf (1901Middeldorf ( -1983, used the term Photothek for the first time for what had been until then called the "image collection" of the KHI.

The Inventory Books
Until 1912, the library and the image collection were not physically separated. Heinrich Brockhaus and his wife Else personally inventoried and cataloged books and images; they were assisted in this by Felix Becker (1864-1928), who had studied under Brockhaus in Leipzig and who would later co-found the Thieme-Becker encyclopedia Allgemeines Lexikon der Bildenden Künstler von der Antike bis zur Gegenwart (General Dictionary of Artists from the Antiquity to the Present).
The Inventarbuch der Abbildungen (inventory book of images) number I was kept as an accession register on the first twenty-two pages from 1898 until 1906. Apart from the fact that the entries were not created at the same time when the images arrived, but rather later on, the inserted sheets with concordances between inventory numbers and the accession register were unclear as well, as were the references in the right column of the inventory book to Abbildungs-Werke (reproductive works), Ansichten (views) and Bildnisse (portraits) (each followed by a Roman numeral). It must be assumed that there were other inventory books besides the "inventory of images." One of those, the volume of "reproductive works," kept among the inventories of the library, has, in fact, been identified as a reference inventory of the "accession register." 28 The entries in it discontinue midway through the "inventory of images" and the last entry was recorded on May 8, 1914. The parallel keeping of the four inventory books thus seems to have been abandoned at that time. To date, the two inventories of "views" and "portraits" mentioned by Brockhaus have not been located. Additional evidence of their existence, however, is provided by an inventory of the furniture and equipment, dated October 1, 1912, 29 and a photograph that was taken between 1912 and 1914 ( Figure 2). This photograph shows on the right, the librarian, Otto Wenzel, and on the left, looking into the camera, Walter Biehl, the Institute's assistant who, like the fellows before and after him, worked part-time at the library and the Photothek. The photograph shows both men in the library after the Institute had relocated to the Palazzo Guadagni on Piazza Santo Spirito in 1912. To the right of Biehl's desk, are photographs and books in a pile-which included Jacob Burckhardt, Der Cicerone (1869)-and on the lower shelf is the inventory of the "views." The title, size, and writing are identical to those of the inventory of the Abbildungs-Werke (Figure 3).
The first "inventory of images" already made use of a cataloging system that remained in effect until the introduction of electronic data entry for new acquisitions  in 1994. Many of the images were later "eliminated as duplicates" and replaced by more recent photographs. Entered into the inventory were the date, the inventory number, the object/subject, the site, the shelf-mark, the producer (of the photograph), the supplier, as well as the price and quantity. The "technique" column initially served to record the technique or medium of the depicted object (drawing, bronze, etc.); later on, the imaging technique (such as zincograph, heliotype, or photography) was entered here.
The first entry in the "inventory of images," which was, as mentioned above, initially kept as an accession register, is dated March 5, 1898, and reads: "Reproductions of artworks, from the estate of Dr. Ulmann. The entire collection (2,959 photographs and 482 halftone prints) has for the time being been loaned to the Institute." The individual images are listed on page 23 of the "inventory of the images," under the date March 5, 1895, and headed by the year 1898. The date of March 5, 1895, is maintained until inventory number 1104, probably because the Ulmann collection was already on loan to the Institute in 1895. 30 The images from this collection are listed in alphabetical order by the artists' names-first painters and then sculptors-up to inventory number 2,713 along with a shelf-mark consisting of the art form and the century of origin.

The Classification System until 1912/1914
In the 1905 annual report, the director describes for the first time the classification scheme for the image collection: "The image collection currently includes roughly 13,700 items compared to 12,900 items the previous year. It is primarily divided into three sections: individual images, reproductive works and coats of arms, while through donations a small foundation has been laid for two additional sectionsportraits and views of cities." 31 The collection of coats of arms consisted of drawings of more than 2,500 coats of arms of "Florentine families, followed by several hundred Florentine churches and brotherhoods, with a systematic index facilitating searches for particular coats of arms; accordingly, there is a chance that it will be possible to trace some Florentine paintings that currently are scattered all over the world back to their original donors by reference to the coats of arms on them." 32 The collection was inventoried in the accession register under the number 35 on January 30,1901, and transferred to the library under the shelf-mark F 9194 where it is still kept today. 33 A more detailed summary of the early arrangement of the books and images of the KHI is also provided by a Führer durch die Bibliothek und Abbildungs-Sammlung (Guide to the Library and Image Collection), which Brockhaus prepared in 1910 ( Figure 4). 34 At the time, the image collection included about 9,000 photographs and some 25,700 prints. These were arranged in four sections: I. The "images" were mainly photographs, arranged by art form (architecture, painting, sculpture), then by country (predominantly Italy), within the country by period, alphabetically first by artist or, if unknown, by site. Kept in envelopes, they were available in two sizes that were placed on top of one another on the same shelf in order for both to be at hand at the same time. Unlike in today's arrangement, the photographs were stored horizontally, rather than vertically in boxes, and they were not distinguished by size.
II. "Reproductive works." These were either printed and bound volumes standing in between the books of the library, or individual sheets that were filed with the photographs in section I. The individual prints were kept separate for conservation reasons, as they were to be stored lying down, rather than standing up like the books.
III. This section, "portraits," covered not just all kinds of printing and photographic techniques, but also all art forms, making it particularly important for research on persons.
IV. Preserved in this fourth section were the "views of cities," which, with a mere 300 images, were likewise still in a nascent stage. At the time this section included only views of Florence and Rome. It also covered all art forms both in terms of imaging techniques and of the images' subjects. Among the documentary materials assembled under the heading "views of cities" were town maps, newspaper clippings, as well as graphic works and other images that went far beyond the scope of a photographic collection.
The classification system in these four sections thus distinguished items not only by the reproduction technology, but also by the depicted objects. Within the sections, a system of shelf-marks was assigned, which were determined exclusively by the depicted objects, highlighting their techniques, dates of origin, and iconographic aspects. The cross-references between the sections that were expressed in the form of Roman and Arabic numerals are difficult to comprehend today because of the missing inventory books ("portraits" and "city views") and the many eliminated images. They often refer to the preceding accession register in which the Klassischer Skulpturenschatz, for instance, was recorded and each individual item was numbered. Probably in 1914, the change was made to systemize with consecutive numbering, adding to the sequence of the numbers of the "inventory of images," which has been continued to this day and by now has reached more than 610,000 inventoried images.
Adding to the difficulty of keeping track of the wealth of images was the fact that different formats were kept in close spatial proximity, even though they were in different envelopes. The art historical classification was noted on the cardboard mount and in the inventory, yet for the purpose of locating and especially for that of scholarly discourse and the method of comparative analysis (Vergleichendes Sehen), it was helpful to put the images together. Since the sections were cataloged in different inventories, creating an adequate finding aid, such as a card catalog, was very important. Moreover, images and books were to be simultaneously accessible to the users of the Institute. This close connection is also reflected in the numerous references to literature worth consulting that were included in Brockhaus's 1910 description of the image collection.

The Shelf-Marks
Although the shelf-marks did draw on the art forms of "painting," "sculpture," and "architecture," the chronological classification was determined by the indicated period (the century): for example, "Ma XV" for painting (Malerei) of the fifteenth century, "Z XVI" for a sixteenth-century drawing (Zeichnung), or "Fr.Ar.XVI" for French architecture of the sixteenth century. In addition, the first photographs related to the applied arts (Kg. XV, i.e., fifteenth-century Kunstgewerbe) were already inventoried on January 30, 1899 (inventory numbers 2754-2756).
By 1912 or 1914 at the latest, this classification by centuries was abandoned in favor of period designations such as "Romanesque," "Gothic," "Renaissance," "Baroque," and "classicism"-epochenbegriffe propagated in German art historical scholarship primarily by Jacob Burckhardt (1818-1897)-and the older photographs and prints that had already been cataloged were given new shelf-marks.
One of the advantages of this reorganization was that works could be bracketed together in larger groups across centuries, as could artists who were active in different centuries. The shelf-mark "Greek sculpture" was subsumed under "antiquity," as Roman copies of Greek originals would otherwise have been found in different sections.
Of course, this reorganization was bound to give rise to other problems. As Reinhart Koselleck noted in his essay in Epochenschwelle und Epochenbewußtsein: "There is something awkward about any division into periods. Many discrepancies have to be accepted, because the suggested divisions cannot be made to align with all the historical evidence. As a result problems arise that, depending on the changing line of questioning, seem to mutually reinforce themselves." 35 Karlheinz Stierle ventured an even stronger statement in his opening sentence in the same volume: "There are no periods. Periods are a way of viewing historical meaning that arranges historical time into figures of interpreted history." 36 It is thus not surprising that one of the early images inventoried-as the first entry after the Ulmann estate in the "inventory of images"-was a print that divided the history of architecture into periods. This color print showing a "Diagram of the development of the periods and stages of the Renaissance style in France from 1475 until 1895" had been created in 1898 by Heinrich von Geymüller (1839-1909) for the Handbuch der Architektur ( Figure 5). Von Geymüller, a Swiss architecture and art historian, who was a friend of Jacob Burckhardt's, was a sponsor of the Institute and a member of the association: "Among the first to visit the newly founded Institute was, in the spring of 1898, Baron Dr. Heinrich von Geymüller who is now also deceased; the last time we welcomed him here again was in November of 1907." 37 In the "inventory of images," under the number 2715, we find not just a detailed entry on this highly complex and intricate diagram of the various styles, but also the note "Framed and mounted," which underscores how much interest was paid to the subject of stylistic periods and its visualization.
One photograph, taken sometime between 1912 and 1914 at the Palazzo Guadagni, shows director Von der Gabelentz-Linsingen in his office where the image collection was kept in boxes, arranged by art forms, and in a uniform size: architecture, sculpture, and applied arts, and, partially visible on the right, the shelves with the boxes marked Malerei (Painting) (Figure 6).
The "architecture" collection included eighteen boxes and was divided into "classical," "early Christian and Byzantine," "Romanesque," "Gothic," "Renaissance," and "Baroque." This was followed by "sculpture" with forty-one boxes, arranged by periods, next by the locations of anonymous works, and finally by artists in the alphabetical order of their names. The shelf ends with several boxes (about ten) covering the applied arts, with their arrangement being based on materials such as ivory, glass, wood, or ceramics.

The Card Index
In his 1910 account, Brockhaus refers to "three card catalogs, divided into the categories 'artists,' 'sites' and 'objects.'" This division is consistent with the Photothek's existing card catalog. In the artist catalog, photographs of attributed works of art are filed separately by art form (painting, sculpture, graphic art, architecture, applied arts). In the site catalog, those photographs-as well as images of anonymous works of art-can be found based on the location of the artwork. And in the iconographical catalog, the objects are arranged according to images, subjects, figures, etc.-to the extent that the visual object allows. Work on the card catalog started as early as 1901. 38 However, ten years later, that work was still incomplete, as we learn from a letter from 1925 by Curt Weigelt, assistant of the KHI 1924-1934, to Wilhelm von Bode: Upon commencing our activity at the Institute we came upon preparatory work for a catalog of our collection of photographs. Yet this work was never finished, but rather … had been abandoned for lack of qualified manpower. And because the use of the photographic device is impossible without a proper card catalog, we have decided to redo this work from scratch as well … . They [the cards] need to be written by typewriter for two reasons: to ensure utter legibility and because the typewriter allows us to create several completely identical copies of each card. 39 What has survived from the original card catalog are mainly those index cards that were "eliminated as duplicates" together with the images, which according to the few date stamps in the "inventory of images" happened between 1924 and 1928. Now useless, those index cards were then relabeled on the back by Ulrich Middeldorf-who had come to the Institute on a fellowship in 1926 and continued to work in the collection until 1935-and reused to compile an "art historically organized catalog." 40 The information about the images themselves was transferred onto the index cards: The labeling of the cardboard supports on which the photographs were mounted, initially corresponded to that of the "reproductive works," such as the Klassischer Bilderschatz or similar reproductive graphics, as well as to captions on photographs. They received an inventory number (top left), a shelf-mark (top right), and a stamp of the KHI with the Florentine fleur-de-lis. This example shows that the cardboard supports were cut to a uniform size at the time they were shelved in boxes in 1912 (Figure 7).
In the first years, the shelf-marks were handwritten in pencil in the inventory book and on the photo cardboard mounts. One could infer from this that the classification scheme was not yet firmly decided on and kept flexible in this way. Later the pencil shelf-marks were replaced by blue stamps, as can still be seen in many examples.
From 1912 onward, the cardboard supports were then labeled in the manner still used today. On the left, below the photograph: the artist and the object depicted; on the right, below the photograph: the site or museum; on the left, at the bottom edge of the cardboard: bibliographical references; and on the right: information about the photograph and the photographer or copyright.

Ulrich Middeldorf and the Expansion of the Classification System since 1926
After completing his studies in Gießen and Munich and doing his doctorate under Professor Adolph Goldschmidt  in Berlin, the twenty-five-year-old Ulrich Middeldorf initially came to the KHI on a fellowship. 41 He had an immense impact on the growth of the holdings, not just because of his personal donations, but also through his contact with one of the most important patrons of the collection, Luigi Vittorio Fossati Bellani, with whose help he systematically expanded sections of the collection. He re-inventoried discarded images and those that had not yet been processed. In addition, he devoted himself to the Institute's own negatives, thereby significantly adding to the collection. 42 Toward the end of these activities, Middeldorf focused more on the applied arts and differentiated relevant objects by materials such as ceramic, leather, porcelain, majolica, silver, and mother-of-pearl, or by techniques, such as inlay, stone carving, and scagliola. On top of that in the period from 1926 to 1935, he donated more than 3,315 photographs to the Institute, a number that was exceeded many times over during his tenure as director from 1953 until 1968. Middeldorf had a particular passion for medals of the Renaissance, which he also collected himself, adding photographs of those objects to the image collection. 43 The inventories, the information about the subject of the images on their cardboard supports, and the changes made to the typewritten index cards of the card catalog all reveal Middeldorf's handwriting, indicating that he applied himself not just to the scholarly classification of the photographs, but also to routine work, such as labeling. 44 Information provided about the work of art or architecture and on classification criteria, such as size and shelf-mark, was now made much more precise. In each of the sections, Middeldorf introduced new sub-classifications. In the architecture section he subsumed views of cities, streets, squares, bridges, etc., under the shelf-mark "city architecture," thereby largely liquidating the section "views." 45 In addition to a revision of the large sizes, the reproductive works (i.e., engravings and other prints) were integrated into the photographic holdings, which led to a commingling of techniques, yet promised more efficient results with regard to image searches.
In the 1929/1930 annual report, Heinrich Bodmer noted: "During the past year the alphabetical and topographical catalogs have undergone a complete revision and also a partial reorganization. The sections Architecture, Sculpture and Applied Arts have been completely reorganized, whereas the Painting section has been partially revised. The compilation of the iconographic index, which had been missing for a long time, has finally begun." 46 Until today, the iconographic index represents an indispensable working tool for detailed research into motifs and themes, such as hagiography, mythology, or history. But, like the card index, the analog means of retrieval, and the inventories, which can be consulted in the Photothek only onsite, they were discontinued after 1993. These were to be replaced by an electronic cataloging system and, in the case of the iconographical card catalog, specifically by means of the Iconclass classification system.
Of particular interest is Middeldorf's discontinued project of an "art historically organized catalog," in which artists were to be cataloged with biographical data, information on their sphere of activity and their teachers, as well as bibliographical references that allowed for a closer correlation with the library. It is not known why this project, compiled by Middeldorf in the 1930s, was abandoned, but most likely it was discontinued after he left the Institute in 1935. 47 Still extant today are two drawers with index cards for painters and one drawer for sculptors. The indexes reveal an arrangement by regions of Italy, next by cities, and within these by period. What was, of course, particularly problematic was assigning each artist to a particular place. Accordingly, there are numerous cross-references to other cities or regions. Moreover, many index cards are filed as "not locatable," which detracts from and compromises the catalog as a finding aid, especially since the index cards lacked references to photographs of works by a particular artist.
Equally instructive-and the only partly surviving-is the catalog with bibliographical references, exclusively to the artists' biographies written by Filippo Baldinucci. 48 An edition of this six-volume work from 1820 found its way into the library of the Institute in 1932, while Middeldorf was working there. Middeldorf returned to Baldinucci again later on and published an article on Baldinucci's Vocabolario toscano dell'arte del disegno in the Institute's own journal. 49 From the Image Collection to the Photothek One of the objectives of the founders of the KHI was the creation and expansion of a comprehensive reference library and image collection on Italian art. From the outset, it was planned as an institutional collection with the ambition to compile the most extensive possible-and thus more or less universal-holdings on Italian art history. In this respect it is different from a private collection of a single scholar that develops specialties, such as the collections of Jacob Burckhardt or Aby Warburg. 50 Yet the first donations and estates, in fact, mostly traced back to private collections and, as a result, certain specialties formed both in terms of the depicted artworks and works of architecture and in terms of media, which necessitated revisions of the classification systems.
Specific purchases could only be made after the funding of the Institute was stabilized-through the support of the German Reich and through the growing membership of the association and its donations. Brockhaus thus writes in the 1904/1905 annual report: "Recently acquired were, specifically, exquisite carbon prints made by Braun in Dornach from paintings in private collections in England that are rarely seen, as well as similarly state-of-the-art large photographs by Alinari, Anderson and Brogi of outstanding artworks in Florence, Rome, Assisi and Padua (Giotto, painting and sculpture of the fifteenth and sixteenth centuries)." 51 Twenty years later, however, the collection had still not grown sufficiently, as Heinrich Bodmer, director of the Institute from 1922 until 1932, pointed out in the first years of his tenure: "Although I donated many thousands of photographs to the Institute this winter, I find that there still are such major gaps that we cannot yet relent in our efforts in this regard; over the course of the next year I hope to get to the point where our photographic device is able to satisfy the reasonable demands of the expert scholars." 52 Bodmer eventually implemented a paradigm shift in the collecting policy that was attributable not just to the 1913 founding of the Bibliotheca Hertziana in Rome, but also to a concentration of the research interests already established at the Institute itself: The campagna fotografica in Bologna and several other experiences I have gained in recent months have made me realize to what extent it corresponds to the specific function of an institute based in Florence that initially the art of Tuscany and, wherever possible, the still largely unexamined art of Florence be made the focus of scholarly research. It is therefore my firm desire to confine myself more and more to the thirteenth to fifteenth centuries and drop all those areas of research that take us too far off from our natural focus, i.e., Florence. With this in mind, I have motivated all the gentlemen affiliated with our Institute to focus their studies on Florentine art, notably of the aforementioned centuries-an activity that can dovetail beautifully with the Institute's planned project of a scholarly guide to the churches. 53 The period that followed thus saw the acquisition and creation of photographs of less developed and more distant marginal fields of art history. On February 16, 1930, for instance, Middeldorf inventoried under no. 87829 a feather picture from the collection of the Palazzo Pitti that was originally from Mexico. It is the first image related to this object that found its way into the Photothek. Over the decades this area has not much expanded, but recently the current director, Gerhard Wolf, has devoted an exhibition in Mexico to the subject of "feather art." 54 The section of "Cimelia," which for decades included only a few images particularly worth protecting and was kept as a "special collection" within the Photothek, has grown through the focus of research on historical photographs. 55 This process of growth and change continues to this day, and the holdings reflect both the interests of those working at the Institute and a certain continuity in the research topics. Newly added research areas and the expansion of the geo-cultural region thus also require an expansion of the classification system and in this sense a collection such as that of the Photothek must be seen as a work in progress.