Berenson did not attend the opening of the Walters Art Gallery on February 3, 1909. Around that time, he was busy cataloging the collections of John G. Johnson and Peter Widener and courting Benjamin Altman as a potential new client. When Berenson met Altman in January 1909, he was buoyed by the prospects of a very lucrative relationship. Altman initially told Berenson, “I want you to make me the finest collections of pictures in the world and make your fortune out of it too. I don’t care what I spend nor how much you make.” Berenson’s ego and the vision of dollar signs that danced through his mind were promptly deflated by Altman’s subsequent observation. Minimizing Berenson’s intellect, Altman told Berenson that he had “the makings of one of the best merchants he ever saw.”1 Stung by this backhanded compliment, Berenson privately denounced Altman a “joker” and a “creature of habit.”2 Then, as the relationship between Berenson and Altman soured and the prospect for obtaining Altman’s business faded, Berenson turned to Walters.
In February 1909, when Walters opened his new gallery, he and Berenson had never met one another. Walters, of course, knew Berenson through his writings and reputation. Similarly, Berenson knew of Walters as a result of his purchase of the Massarenti collection. In Northern Italian Painters of the Renaissance, which Berenson published in 1907, six of the paintings that Walters had purchased from Massarenti were listed. Each painting was listed as belonging to “Baltimore, U.S.A. Mr. Henry Walters.”3
On July 8, 1909, Berenson, seeking to attract Walters as a client, contacted a mutual acquaintance, William Laffan, and asked him to convey to Walters a proposal to prepare a catalogue of Walters’s recently displayed collection of Italian paintings. Laffan informed Berenson that he thought that the offer was attractive, and he would recommend it to Walters.4 On July 28, 1909, Laffan informed Berenson that Walters was impressed with his offer. He also conveyed that Walters was not interested in publishing a run-of-the-mill catalogue; Walters wanted something much more significant that would take the form of a “classical edition.”5 To further entice Walters to engage him to write his catalogue, Berenson, in August 1909, informed Laffan that he would not charge any money to do this.6
Walters did not jump at Berenson’s offer. He probably was worried that Berenson already had acquired a negative impression of the Massarenti collection and that Berenson would be inclined to critically evaluate the paintings with icy derision. There was good reason for this apprehension. When Berenson wrote his Four Gospels on Venetian, Florentine, Northern Italian, and Central Italian painters, he was well aware of the Massarenti collection. But he gave no credit to Massarenti’s claim that the collection included paintings by Michelangelo, Raphael, Titian, and other Italian masters. In The Venetian Painters of the Renaissance, published in 1894, Berenson referred to the Massarenti collection as having no significant Venetian paintings other than those by Guardi and Polidoro. In The Central Italian Painters of the Renaissance, Berenson referred to the Massarenti collection as having paintings only by three relatively undistinguished artists, Mariotto, Pacchiarotto, and Zagnelli. Moreover, in The Florentine Painters of the Renaissance, Berenson did not refer to any masterpieces in the Massarenti collection. Because Walters had adopted most of Massarenti’s attributions, there was reason for him to apprehend that Berenson would discount his attributions as quickly as he had dismissed Massarenti’s. Walters’s reluctance to engage Berenson to examine his paintings, however, was overcome by the prospect of having the world’s greatest authority on Italian Renaissance art prepare a catalogue for the Walters collection and thereby give it his invaluable stamp of approval. It was a gamble as risky as the acquisition of the Massarenti collection itself.
On September 16, 1909, Walters wrote to Berenson that, having reviewed Berenson’s offer regarding his catalogue, he was “pleased at the prospect of having you undertake this for me.” As reflected in this letter, Walters at that time envisioned that a catalogue of his Italian paintings endorsed by Berenson would be in great demand. He stated that it was his intention “to issue a very large edition of the catalogue,” that he expected the catalogue to include reproductions of “every Italian picture” in his collection except for some minor ones, and that he planned to distribute the catalogue to “all libraries of importance.” In addition, he wanted to sell the catalogue at a nominal price at his gallery. He also hoped to publish a second, special edition of two hundred copies for private distribution, half of which he would keep and the other half he would give to Berenson. He concluded his letter with a small dose of reality. Anticipating Berenson criticism of some if not many of his paintings, Walters wrote, “I have no doubt that there are now pictures on my walls which will be removed permanently or replaced by others more worthy.” In this regard, Walters deferred to Berenson, writing that “I will be governed by your suggestions.”7
The prospect of publishing a beautiful catalogue of his Italian paintings written by Berenson must have been extremely appealing to Walters not only because it would have elevated his own status in the art world but also because the publication of beautiful books that showed off the Walters art collection had been an important part of his father’s cultural legacy. During his lifetime, William Walters published a series of elaborate, richly illustrated leather-bound books about his collection, including two black leather-bound editions of the Collection of W. T. Walters and the Oriental Collection of W. T. Walters, both published in 1884. In 1897, three years after William Walters’s death, a luxurious large folio entitled Oriental Ceramic Art: Illustrated with Examples in the Collection of W. T. Walters was published. It contained 116 large, beautifully colored lithographs mounted between double sheets of silky Oriental paper and included a preface written by William Laffan. Henry Walters distributed a limited edition of this beautiful book to various museums and libraries in the United States and abroad.8 Walters probably had this book in mind when discussing with Berenson his plan to publish worldwide his own book of Italian paintings.
In 1909, at the time Berenson committed to producing an important, scholarly catalogue for Walters of his collection of Italian paintings, Berenson already was working on a similarly ambitious catalogue for his client John G. Johnson. The Johnson catalogue was an important and time-consuming project. Berenson characterized it as “a labour of love.”9 He periodically worked on the catalogue for five years, from 1908 to 1913. By the time of its completion, the Johnson catalogue was more than three hundred pages long and contained descriptions of over two hundred paintings, thumbnail sketches of their artists, and more than two hundred illustrations.
Walters was well aware that Berenson’s work on his catalogue would have to await the completion of the Johnson catalogue. He informed Berenson that “[I] only regret that it will not be possible to have . . . [my] catalogue ready at an earlier date.”10 When Berenson showed Walters a draft of the Johnson catalogue, Walters, according to Berenson, was “delighted” with it, and it “aroused his envy.”11 At that time, both Berenson and Walters contemplated that Walters’s catalogue would be like Johnson’s, that the two catalogues would be connected pedagogically, and that together they would constitute the cornerstones of any study of Renaissance art in America. Berenson expressed this idea to Johnson, stating that he hoped that both catalogues when considered together would provide American students of the Renaissance with a “pretty adequate idea of the more formal side at least of Italian painting.”12 Berenson’s idea made sense. The Johnson and Walters collections of Italian paintings were related not only by the fact that they contained many of the same artists but also by the common interest of their owners in using their collections to relate the historical development of Italian Renaissance painting. Among the thirty-four paintings that Berenson would sell to Walters, nineteen were by artists who were already in the Johnson collection.13
As had been done in preparing the Johnson catalogue, Berenson requested that Walters provide him with photographs of all of the Italian paintings listed in Walters’s 1909 catalogue. In response to Berenson’s request, Walters, in October 1909, instructed his curator, Faris Pitt, to send the photographs to Berenson as soon as possible in order for Berenson “to prepare the catalogue.”14 In November 1909, Pitt sent 286 photographs, 191 of which were silver prints and 95 of which were platinum prints, to Berenson. Walters noted on the back of each photograph the identity of the artist or the school to which the painting was attributed. The photographs were accompanied by a letter dated November 13, 1909, in which Walters reiterated his request for Berenson’s advice while offering the self-serving suggestion that he had already begun distinguishing the good pictures from the bad: “I do wish you would indicate what pictures you think should be withdrawn definitely from the collection. I myself have already in my mind made eliminations.”15
Having invited Berenson to advise him about what paintings to keep and what to discard, Walters attempted to discourage him from cutting too deeply into the collection. As he had done with the exaggerated story about leaving half of the Massarenti collection in Rome, which Walters had invented to reduce the size of his import tax, Walters conveyed to Berenson that he had already disposed of most of the second-rate paintings in his collection. On January 4, 1910, he wrote to Berenson that he had already discarded “over four hundred and fifty pictures, Italian, French, Dutch and non-descript, and I am quite prepared to add to the rejected one hundred more.” There is no evidence in the Walters archives to support this fanciful assertion. Indeed, Walters did not begin to dispose of any of the paintings he had acquired from Massarenti until 1915, approximately five years later. Walters also informed Berenson that he had fifteen Italian paintings, including Raphael’s Madonna of the Candelabra, which he had purchased independently of the Massarenti collection (see appendix B, list 3). With regard to the painting by Raphael, Walters stated that it was the “most important” individual painting he had acquired. In an effort to preempt any second-guessing by Berenson about the painting’s provenance, Walters emphasized that he had carefully traced its history: “The most important one being the Madonna of the Candelabra by Raphael, which I bought from a nephew of Monro, of Novar, it having been taken from the Borghese Collection by Napoleon I and given to the Queen of Etruria, from whom it passed to the Duke of Lucca, who sold it at Christies in London, where it was bought by Mr. Monro.”16 (A copy of this entire letter is reprinted in appendix A.) It is doubtful that Berenson gave much weight to Walters’s recital of the painting’s provenance. As he would later state to Duveen, “You must never encourage any question of provenance. In most cases . . . it is impossible to know, and at other times impossible to tell.”17
Berenson was dissatisfied with the quality of some of the photographs sent by Walters and requested additional photographs of more than one hundred paintings. Still believing that this effort would result in a magnificent catalogue authored by Berenson and illustrated with Walters’s collection of Italian paintings, Walters replied, “I am anxious to supply you with what you require.”18 By the end of March 1910, Walters had provided Berenson with at least one photograph of every Italian painting in his collection.
In the spring of 1910, Berenson invited Walters to be his guest at I Tatti. Although they had become acquainted through their correspondence, they still had not met face to face. In response to Berenson’s invitation, Walters wrote, “I am very anxious to have the pleasure of a personal acquaintance.”19 As evidenced by their subsequent correspondence, Walters and Berenson discovered that they shared a variety of mutual interests that bonded their relationship and spawned a genuine friendship. Their letters were not only about the hard, dollars-and-cents business of selling and buying paintings but also addressed personal issues related to their health, their families, their homes, their economic and political philosophies, and current events on both sides of the Atlantic. Although of dissimilar age (in January 1910, Walters was sixty-two and Berenson was fifty-five), they commiserated about the frailties associated with growing old. Berenson’s concerns were psychological, with veiled references to suicide; Walters suffered from problems with his gallbladder. They shared similar, conservative political philosophies, both vowing allegiance to the Republican Party and supporting the presidency of Theodore Roosevelt. They both expressed apprehension over the growth of communism, disappointment over America’s neutrality in the face of German hostility, and concern about the effects of the war on their business interests. They shared a zest for the finer, material pleasures of life that their fame and fortune had brought abundantly to both of them. And an overarching devotion to high culture was their common religion.
During his visit to Berenson’s villa in May 1910, Walters was impressed with the quality of the Renaissance paintings that were available for sale. The overall cost of purchasing such paintings from Italy had declined the previous year as a result of the Payne Aldrich Tariff Bill, which had abolished import duties on works of art that were more than one hundred years old. Mindful of this reduction, Walters inquired about the cost of the paintings that Berenson had available for sale. Upon receiving this information, on May 23, 1910, Walters sent Berenson a note stating that he was “greatly delighted with the result” and that he wanted to arrange the “methods of payment at present and in the future.”20 As Berenson had planned, this led to an agreement by Walters to engage Berenson to acquire paintings for him.
The precise terms of this agreement were memorialized in a letter from Walters to Berenson dated May 24, 1910:
In accordance with our understanding, it will be most interesting for me to place at your disposal between July 1, 1910 and July 1, 1911, the sum of Seventy Five Thousand Dollars ($75,000), or so much thereof as you may require during that period, to pay for purchases for my collection of pictures, particularly of the Italian schools, in order to fill out its historical value and increase its average quality . . . To give this arrangement a proper business relation, you are to receive as a commission from me upon each purchase made and at the time of each purchase ten percent of its cost.21
(A copy of this entire letter is reprinted in appendix A.) As further detailed in this letter, Berenson was required to send photographs of the paintings he offered to Walters, and Walters would respond in writing as to whether he accepted the offer. To assure that the paintings he purchased from Berenson would be exempt from any import duty under the Payne Aldrich Tariff Bill, Walters’s letter also confirmed that Berenson would identify himself as the seller of each painting and provide a written guarantee that “the object is a work of art and more than one hundred years old.”22
Following his acquisition of the Massarenti collection and prior to retaining Berenson as his principal dealer, Walters had acquired fifteen Italian paintings (see appendix B, list 3). Perhaps worried about the efforts in Italy to prevent the loss of its cultural treasures,23 Walters approached the acquisition of these paintings as if on a secret mission. With only one exception, he retained no formal records relating to the source of these paintings, the cost of the paintings, the date of their acquisition, or the manner in which they were shipped to Baltimore. In his contract with Berenson, Walters expressed his continuing concern for secrecy. He did not want to be directly linked in a financial relationship with the original Italian seller, and he directed Berenson to serve as a buffer between them. In this regard, the terms of Walters’s arrangement with Berenson provided: “I prefer to have all settlements with you directly so that my name will not except in exceptional cases appear to the seller. In other words so far as I am concerned in the settlements you pay out this money for me.”24
Although the subject matter was high culture, the agreement was clearly intended to establish a formal business relationship. Walters made this explicitly clear by reducing the agreement to writing, by using phrases that carried an air of legality, and by expressly stating that their agreement should be treated as “a proper business relationship.” The agreement contemplated but did not bind either side to a long-term relationship. Under its terms, the agreement lasted for one year, from July 1, 1910, to July 1, 1911, leaving open the possibility of annually renewing it or permitting their business relationship to be thereafter terminated at the will of either party at any time. The terms of their agreement also did not contractually bind Berenson and Walters together in any exclusive relationship. Berenson remained an independent contractor who was free to sell paintings to other collectors like John G. Johnson. Walters likewise remained free to purchase Italian paintings from anyone he chose, and he continued to purchase Italian paintings from many other sources, including A. S. Drey, a prominent dealer in Munich and New York. To effectuate the sale of paintings to Walters, Berenson agreed to ship the paintings to Maurice Pottier, Walters’s shipping agent in Paris, who would forward the paintings to Walters in Baltimore.
The financial terms of the contract provided Berenson with the opportunity to earn a comfortable income that could continue for many years. If, as contemplated by the terms of the agreement, Walters spent $75,000 each year on paintings purchased from him, such expenditure would generate $7,500 annually in commissions to Berenson. By today’s standards, this was equivalent to approximately $165,000 each year.25 Although unmentioned in the contract, Berenson was eligible to receive additional compensation in the form of an “honorarium” for reviewing and approving paintings that Walters purchased from other dealers. For example, in March 1914, Walters sent Berenson a check in the amount of $3,000 for examining paintings that Walters acquired from the A. S. Drey Gallery.26 Retaining the freedom to engage in business with other collectors as well as dealers, Berenson’s financial arrangement with Walters must have been very attractive to him.27
A critical provision in the agreement between Walters and Berenson involved the standard by which Berenson would select and Walters would measure the quality of the paintings that were to be added to Walters’s collection. The standard used in the agreement was to acquire paintings that would “increase its [the collection’s] average quality.” What is striking about this standard is how modest it was. Walters was not in the market for the world’s greatest Italian masterpieces. He simply wanted good paintings that, in general, were better than what he already had and that would enhance the educational value of the collection. The second unusual aspect of this standard was the difficulty of its application. This standard was not fixed or objective but soft, subjective, and comparative in nature. It related to the existing quality of Walters’s collection. It implicitly required Berenson to review and evaluate the overall quality of Walters’s existing collection in order to intelligently select paintings that were better than the existing collection’s “average quality” and that improved its capacity to fairly represent the history of Italian art. Because Berenson and Walters had sharply different opinions about the quality of Walters’s collection, the standard employed in their agreement was difficult to measure. As reflected in his 1909 catalogue, Walters at that time optimistically expressed the belief that artists in his collection included many of the masters of Renaissance art and that the collections was, in general, important enough to warrant the publication of a widely disseminated catalogue. On the other hand, Berenson, having reviewed Walters’s photographs of his Italian paintings and having recognized the presence of many rank copies, perceived the “average quality” of the collection to be appreciably lower. From Berenson’s viewpoint, improving the average quality of Walters’s collection was an achievable standard that was easy to meet.
Although Walters, as a general rule, was very secretive and averse to publicity, he conveyed to the press that Bernard Berenson had agreed to inspect Walters’s paintings and help him prepare a catalogue.28 He recognized that the mere association of Berenson’s name with his collection gave it a panache and credibility that was otherwise unattainable. On November 6, 1910, the Baltimore News American reported this story under this headline: “Bernard Berenson, Eminent Italian Critic, Made Careful Inspection of Henry Walters’ Massarenti Collection.” The column added that, based on the Walters-Berenson collaboration, “A very handsome catalogue . . . will be published.”29