“Three Long Rows of Empty Shelves” to Fill: Curt Nimuendajú as Collector and Researcher for Ethnological Museums in Germany, 1928-1930

In 1928-29 and 1930, the German-Brazilian anthropologist Curt Nimuendajú was contracted twice by German ethnological institutions (above all, museums) for organizing ethnographic collections and carrying out anthropological research among indigenous peoples, principally Jê-speaking peoples in various regions of today’s Maranhão and Tocantins states in Brazil. This is not only a lesser-known part of Nimuendajú’s biography, but also an example of a kind of academic cooperation difficult to imagine nowadays. The collections, partly destroyed during World War II, are still stored in the ethnological museums of Hamburg, Leipzig, and Dresden, along with a great number of mostly unpublished letters and other documents linked to them. The history of these expeditions offers an opportunity for reflections about the implicit theories involved in contemporary collecting, fieldwork methods, and the style of anthropology practiced. The research reveals influences from German ethnology on the academic environment in Brazil, which later became increasingly independent from this input. keywords Nimuendajú, German ethnology, Ethnographic Collections, Ethnological Museums, Jêspeaking Peoples DOI http://dx.doi.org/10.11606/ 2179-0892.ra.2019.157040 218 special issue | Peter Schröder | “Three Long Rows of Empty Shelves” to Fill: Curt Nimuendajú as Collector and Researcher for Ethnological Museums in Germany, 1928-1930 Rev. antropol. (São Paulo, Online) | v. 62 n. 1: 217-240 | USP, 2019 introduction Today, the German-Brazilian anthropologist Curt Nimuendajú is considered one of the pioneering figures in the history of Brazilian anthropology in its pre-institutional period. In 1928-29 and 1930, he was contracted by German ethnological institutions for organizing ethnographic collections and carrying out anthropological research among indigenous peoples in various regions of the current Maranhão and Tocantins states. This article is about the partial results of a research project regarding Nimuendajú’s relationships with German museums. As a contribution to the history of Brazilian anthropology and of German ethnology, it deals with the history of two field expeditions in Brazil and the ethnographic collections that resulted from them.1 I discovered this subject in 2009 by reading a popular biography of Nimuendajú (only two biographies of him exist, both are non-scientific2), published in the former GDR, by a journalist and Heimatforscher (regionalist amateur researcher) in 1979. In his book Nimuendajú – Brother of the Indians, Georg Menchén, who died in 1989, published part of the correspondence between Nimuendajú and his interlocutors in Germany. The main problems with this book are that Menchén’s selection of sources was one-sided, biased, and his narrative is not trustworthy because the author substituted a lack of information with imagined episodes, often with romantic traits. That same year, I initiated e-mail contact with German ethnological museums in order to find out if the collections organized by Nimuendajú still existed and if the documents cited by Menchén could be consulted; the answers were positive. In 2010-11, during a seven-month period of post-doc research in Germany,3 I carried out a survey regarding the collections organized by Nimuendajú for German ethnological museums and the documentation related to them (Schröder, 2011). Research was done, among other place, in the following places and institutions: <right> The Grassi Museum in Leipzig; <right> Archives of the Institute of Ethnology at Leipzig University; <right> The Museum of Ethnology in Dresden; <right> The Museum of Ethnology in Hamburg; <right> The Ethnological Museum in Berlin; <right> The Ethnological Museum in Munich (now Museum Fünf Kontinente). Indeed, in Hamburg, Dresden, and Leipzig, the majority of ethnographic and archeological objects collected by Nimuendajú for German museums can still be found, confirming the information found in two former publications (Becher, 1 This article is a revised and extended version of a paper originally presented in the Panel 060 “Themes in the History of Anthropology and Ethnology in Europe” during the 14th Biennial Conference of the European Association of Social Anthropologists (EASA), held at the University of MilanoBicocca in Milan, Italy, 20-23 July 2016. The presentation of the paper was made possible by a grant from CNPq (Conselho Nacional de Desenvolvimento Científico e Tecnológico) of the Brazilian Ministry of Science, Technology, Innovations and Communications (MCTIC) (AVG facility, process n. 451989/2016-3). The author also expresses gratitude to CNPq for supporting studies about the work and life of Nimuendajú and his contacts with German ethnology by having granted a special scholarship for distinguished productivity called Produtividade em Pesquisa (PQ modality; processes n. 307304/2013-2 [2014-2017] and 306550/2016-4 [2017-2020]). 2 A second biography of Nimuendajú was completed in 2014 in Bochum (Germany) by the Brazilian journalist Salvador Pane Baruja and was made available online in 2015 (Pane Baruja, 2014). Although his treatise, in comparison with Menchén’s book, has the evident merit of being extensively based on Brazilian and German documentary and image sources, his analysis in many parts suffers from a lack of scientific rigor. The author, for example, explains Nimuendajú’s thematic preferences in a certain field research by referring to specific psychological constellations of his biography instead of trying to relate them to prevailing tendencies in German ethnology at that time. 219 special issue | Peter Schröder | “Three Long Rows of Empty Shelves” to Fill: Curt Nimuendajú as Collector and Researcher for Ethnological Museums in Germany, 1928-1930 Rev. antropol. (São Paulo, Online) | v. 62 n. 1: 217-240 | USP, 2019 1955; Staatliches Museum für Völkerkunde Dresden, 1983) about the existence of the collections. The hook for this article is that a detailed evaluation of new information contained in original sources only partially analyzed, allows us to draw some interesting conclusions about the collector; as well as about the historical context of his activities.

introduction Today, the German-Brazilian anthropologist Curt Nimuendajú is considered one of the pioneering figures in the history of Brazilian anthropology in its pre-institutional period. In 1928-29 and1930, he was contracted by German ethnological institutions for organizing ethnographic collections and carrying out anthropological research among indigenous peoples in various regions of the current Maranhão and Tocantins states. This article is about the partial results of a research project regarding Nimuendajú's relationships with German museums. As a contribution to the history of Brazilian anthropology and of German ethnology, it deals with the history of two field expeditions in Brazil and the ethnographic collections that resulted from them. 1 I discovered this subject in 2009 by reading a popular biography of Nimuendajú (only two biographies of him exist, both are non-scientific 2 ), published in the former GDR, by a journalist and Heimatforscher (regionalist amateur researcher) in 1979. In his book Nimuendajú -Brother of the Indians, Georg Menchén, who died in 1989, published part of the correspondence between Nimuendajú and his interlocutors in Germany. The main problems with this book are that Menchén's selection of sources was one-sided, biased, and his narrative is not trustworthy because the author substituted a lack of information with imagined episodes, often with romantic traits.
That same year, I initiated e-mail contact with German ethnological museums in order to find out if the collections organized by Nimuendajú still existed and if the documents cited by Menchén could be consulted; the answers were positive. In 2010-11, during a seven-month period of post-doc research in Germany, 3 I carried out a survey regarding the collections organized by Nimuendajú for German ethnological museums and the documentation related to them (Schröder, 2011). Research was done, among other place, in the following places and institutions: Indeed, in Hamburg, Dresden, and Leipzig, the majority of ethnographic and archeological objects collected by Nimuendajú for German museums can still be found, confirming the information found in two former publications (Becher,1 This article is a revised and extended version of a paper originally presented in the Panel 060 "Themes in the History of Anthropology and Ethnology in Europe" during the 14th Biennial Conference of the European Association of Social Anthropologists (EASA), held at the University of Milano-Bicocca in Milan, Italy, 20-23 July 2016. The presentation of the paper was made possible by a grant from CNPq (Conselho Nacional de Desenvolvimento Científico e Tecnológico) of the Brazilian Ministry of Science, Technology, Innovations and Communications (MCTIC) (AVG facility, process n. 451989/2016-3).
The author also expresses gratitude to CNPq for supporting studies about the work and life of Nimuendajú and his contacts with German ethnology by having granted a special scholarship for distinguished productivity called Produtividade em Pesquisa (PQ modality; processes n. 307304 /2013-2 [2014-2017] and 306550/2016-4 [2017-2020]).

2
A second biography of Nimuendajú was completed in 2014 in Bochum (Germany) by the Brazilian journalist Salvador Pane Baruja and was made available online in 2015 (Pane Baruja, 2014). Although his treatise, in comparison with Menchén's book, has the evident merit of being extensively based on Brazilian and German documentary and image sources, his analysis in many parts suffers from a lack of scientific rigor. The author, for example, explains Nimuendajú's thematic preferences in a certain field research by referring to specific psychological constellations of his biography instead of trying to relate them to prevailing tendencies in German ethnology at that time. special issue | Peter Schröder | "Three Long Rows of Empty Shelves" to Fill: Curt Nimuendajú as Collector andResearcher for Ethnological Museums in Germany, 1928-1930 1955; Staatliches Museum für Völkerkunde Dresden, 1983) about the existence of the collections. The hook for this article is that a detailed evaluation of new information contained in original sources only partially analyzed, allows us to draw some interesting conclusions about the collector; as well as about the historical context of his activities.

the main actors
Among the main actors involved in the story of the two expeditions done for German ethnographic museums -Nimuendajú, Fritz Krause, and Otto Reche -Nimuendajú is certainly the best known in the international anthropological community.
Curt Unckel was born on April 17, 1883, in Jena, Thuringia, Germany and immigrated to Brazil in 1903. From 1905to 1907, he spent two years with a Guarani group in a village along the Batalha River in the hinterland of São Paulo state, where he was adopted by an indigenous family and received his name Nimuendajú. He later registered Nimuendajú as his family name in 1926 when opting for Brazilian citizenship. In 1913 he moved from São Paulo to Belém, where he maintained as his permanent address until his death in December 1945 in a Ticuna village, in São Paulo de Olivença municipality, in the Upper Solimões region of the Amazon (Figure 1). The cause of his death is still unclear, although the hypothesis that he was mudered is the most accepted (Oliveira, 1999). 4 More than four decades of work dedicated to the ethnology of indigenous peoples earned him, already during his lifetime, the reputation of being one of the best experts on indigenous peoples of Brazil in the first half of the century (Grupioni, 1998: 164) and, according to some, the greatest of all (Kraus, 2004: 44-45).
It is interesting to note that Nimuendajú is not only attributed to one national tradition in anthropology; however, this depends on one's point of view. In Brazilian anthropology, he is unequivocally seen as a Brazilian anthropologist, sometimes with the addendum "of German descent," but generally he is not con-4 In a recent article Elena Welper offers a kind of genealogy of the different versions of Nimuendajú's death. See Welper (2016). 3 Financed by CNPq. The scholarship was a PDE facility (process n. 200455/2010-9). Hereby, I again express my gratitude to CNPq for having made possible this research.

Figure 1
Curt Nimuendajú during fieldwork, around 1934 (unknown photographer; Acervo Instituto Martius-Staden, São Paulo) special issue | Peter Schröder | "Three Long Rows of Empty Shelves" to Fill: Curt Nimuendajú as Collector andResearcher for Ethnological Museums in Germany, 1928-1930 sidered a representative of German ethnology. On the other hand, in Germany, things look very different. For example, among the short biographical portraits of the portal Interviews with German Anthropologists one can find an entry about "Curt Nimuendajú (Unckel)." 5 Furthermore, in 2013 I was invited by the Historical Commission of the Bavarian Academy of Sciences to write a dictionary article on him for volume 26 of Neue Deutsche Biographie (New German Biography), an important reference which was finally released in 2016 (Schröder, 2016), though Curt is listed as Unckel, not as Nimuendajú. Fritz Krause (1881-1963) was a German ethnologist principally known among Americanists for his 1908-09 expedition to the Araguaia region in Central Brazil. He found employment at the Ethnological Museum in Leipzig from 1912 onwards, serving as its director since 1927. During this time he also taught as a professor at Leipzig University from 1925 to 1945 ( Figure 2). Krause developed a distinct approach to ethnology, which he labelled "ethnological structuralism." Although having strong psychological associations, it was very different from the later French structuralism of Lévi-Strauss. Krause's intellectual effort was aimed at finding a new theoretical path among the mainstream tendencies of German ethnology at that time (see Wolfradt, 2011 for Krause and his structuralist approach). As a co-founder of the German Ethnological Society (Deutsche Gesellschaft für Völkerkunde -DGV) in 1929, Krause held a strong position in the institutional arrangements of German ethnology at that time, but after World War II his arrangements and engagements with National Socialism gave him the status of a person hardly tolerated in Eastern Germany under the communist regime. Nevertheless, his political sympathies did not appear in his correspondence with Nimuendajú (for an obituary of Krause, see Damm, 1966).
Otto Reche (1879Reche ( -1966, an ethnologist and physical anthropologist, was one of the most prominent exponents of Nazi anthropology. Being a tenure track professor of (physical) anthropology and ethnology at Leipzig University since 1927, he was director of its Institute for Anthropology and Ethnology, later renamed "Institut für Rassen-und Völkerkunde" (Institute for Racial 5 Available at: http://www. germananthropology.com/ short-portrait/curt-nimuendajunckel/193, access in July 13, 2018.

Figure 2
Fritz Krause (Reproduced from Herbert Baldus, Bibliografia Crítica da Etnologia Brasileira, vol. 2., Hannover: Kommissionsverlag Münstermann-Druck, 1968;unknown photographer) special issue | Peter Schröder | "Three Long Rows of Empty Shelves" to Fill: Curt Nimuendajú as Collector andResearcher for Ethnological Museums in Germany, 1928-1930 Studies and Ethnology) (for the academic vita of Reche, see Geisenhainer, 2002). Again, as in the case of Krause, his ideological and political ideas did not appear in his correspondence with Nimuendajú.
Other German scientists were directly and indirectly involved in the history of the two expeditions. Their roles were secondary in comparison to the three main actors mentioned above because they never were direct interlocutors with Nimuendajú, exchanging letters only with Krause or Reche. Among these "secondary" actors we should include Georg Thilenius , director of the Hamburg Museum of Ethnology, and Arnold Jacobi , zoologist and director of the Dresden Museum of Zoology and Ethnology; both were involved in financing the expeditions. Their correspondence with Krause, and on a smaller scale with Reche, above all reveal the financial and logistic preoccupations and less an interest in the circumstances of Nimuendajú's fieldwork. However, this idea may be the result of the letters' subjects, because they received copies of all letters sent by Nimuendajú to Krause from the fieldwork areas.
Thus, the main axis of communication existed between Krause and Nimuendajú. However, there was another direct interlocutor for Nimuendajú, who must be mentioned: Carlos Estevão de Oliveira (1880Oliveira ( -1946 of the Museu Paraense Emilio Goeldi (Figure 3), one of his most important allies in the Brazilian institutional scene and a personal friend for the rest of his life. All the letters from Nimuendajú to Carlos Estevão, held at the State's Museum of Ethnology and Archaeology of São Paulo, have been published by Thekla Hartmann (see Nimuendajú, 2000), and some of them provide accounts of his fieldwork activities for the German museums.

the sources
The primary sources of this research are letters and complementary documents, lists of objects, maps, photos, and ethnographic objects, are all archived and deposited in the ethnological museums of Leipzig, Dresden and, on a minor scale, Hamburg. In the Ethnological Museum at the Grassi Museum Leipzig the entire documentation of both expeditions was found in near entirety. Visual materials are quite scarce, while the number of well conserved correspondence is remarkable 6 .
Other correspondence was also found; principally between Krause and the Swiss ethnologist Felix Speiser , the Swedish ethnologist Baron Erland Nordenskiöld (1877-1932, director of the Ethnographic Museum of Gothenburg, and the Austrian ethnologist Father Wilhelm Koppers (1886Koppers ( -1961, from the Anthropos journal. These letters only represent a minor part of the correspondence regarding the two expeditions.
The documentation archived at the Grassi Museum allows a complete reconstruction of the history of the first expedition. Whereas, in the case of Nimuendajú's second expedition for German museums, such a reconstruction is only possible with the help of the documents found in Dresden. Generally, Nimuendajú's letters are long, detailed, and have minimal spacing. They often contain vivid descriptions of the fieldwork conditions and his research style, together with detailed ethnographic information and numerous critical observations about indigenous politics as well as about the kinds of relationship between the indigenous and non-indigenous population. Frequently, the language of these letters is non-scientific and generally has a mixture of formal German writing and popular expressions. Without exception, all the letters were written in German.
From these letters, a quite different Nimuendajú emerges than the one known by his scientific monographs and articles. By comparison, the language of Krause, Reche, and that of the other German ethnologists, is sober and even bureaucratic, although Krause is the only one among them who sometimes makes ironic comments in some passages, adding a different, more interesting style to his letters.
One example might illustrate this. In January, 1928, Nimuendajú wrote a letter to Krause with two proposals for expeditions which could be conducted by him for German museums 7 : An ethnographic travel for collecting, with the center of activities in Boa Vista on the Tocantins River, with the Apinayé as the principal object for studies and an extension of the studies to the neighboring tribes as far as the means can afford. I consider this field of activities by far the best. Very good results can be expected, because I am quite sure of the Apinayés' good intentions. [...] In Boa Vista, there is a an unrestricted ruler governing, the old Father João, without whose most holy will no one can land firmly on the ground there. Much depends upon his thoughts. As far as I know, at least he is no enemy of the Indians. 6 The following original letters were located: 53 letters exchanged between Nimuendajú and Krause about the two expeditions; 12 letters between Nimuendajú and Reche regarding the second expedition; 10 letters between Krause and Reche regarding the second expedition; 29 letters between Krause and Thilenius about the two expeditions; 28 letters between Krause and Jacobi about the first expedition. In the depot of the Museum of Ethnology in Dresden, the following letters were found, either as originals or as photocopies: 15 letters between Nimuendajú and Krause; 34 letters between Nimuendajú and Reche; and 14 between Krause and Reche.
On a lesser scale, there are also some letters between Krause, Thilenius, and Jacobi, and between Reche, Thilenius, and Gustav Antze, from the Hamburg Museum. The documents archived in Dresden enable a reconstruction of major aspects of the second expedition. As for the style of the letters, the same observations can be made as in the case of the correspondence archived at Leipzig. So far, it was not possible to visit the archive at the Hamburg Museum of Ethnology, but this does not represent a major problem. This is because the entire correspondence between the three museums could be localized in Leipzig and Dresden, and there was no direct contact between Nimuendajú and the Hamburg museum staff.

contact and contract
Nimuendajú was a self-educated anthropologist and never had the possibility of an academic education. He had his first contact with the European academic environment when he published his famous ethnographic monograph about Guarani religion and worldview in the well-respected Zeitschrift für Ethnologie (Nimuendajú, 1914). He achieved this success through the mediation of the German ornithologist Emilie Snethlage , who had become director of the Goeldi Museum in Belém that year. World War I interrupted this contact with German ethnology but after the war Nimuendajú managed to maintain a regular correspondence with Theodor Koch-Grünberg (1872-1924 for nine years, from 1915 to 1924, terminating with the death of Koch-Grünberg. 9 Having published various articles in periodicals such as Zeitschrift für Ethnologie (Berlin), Petermanns Geographische Mitteilungen (Gotha), Anthropos (Mödling bei Wien), and Journal de la Société des Américanistes (Paris), Nimuendajú built up a reputa-8 SES, Leipzig, 1928/43, Krause, April 13, 1928 9 The entire correspondence between Nimuendajú and Koch-Grünberg will be published by the Museu Paraense Emilio Goeldi (MPEG) in a volume edited by Michael Kraus (Göttingen University), Ernst Halbmayer (Marburg University), Nelson Sanjad (MPEG), and the present author. special issue | Peter Schröder | "Three Long Rows of Empty Shelves" to Fill: Curt Nimuendajú as Collector and Researcher for Ethnological Museums in Germany, 1928-1930 tion as a specialist in indigenous ethnology and linguistics of Lowland South America in the midst of the 1920s.
But he was short on money, and this seems to have been a permanent cause for concern until the end of his life. At the very least, this regularly is one of the topics in his letters. For this reason, he accepted to carry out archaeological excavations and ethnographic studies for the Ethnographic Museum of Gothenburg, from 1923Gothenburg, from to 1927, with the support of the museum's director, Erland Nordenskiöld (Nimuendajú, 1929a;2004). But in 1926 the relationship with this sponsor became complicated and the cooperation ended in 1927. Thus, the continuity of Nimuendajú's studies was at risk.
In this delicate situation, a recommendation by the Swiss ethnologist Felix Speiser was very welcome. Speiser was introduced to Nimuendajú through letters from Koch-Grünberg and later met him in Belém. On February 25, 1927, Speiser wrote a letter to Krause, introducing Nimuendajú as an excellent collector and fieldworker: It is noteworthy that Krause's questions were not only about collecting objects in a strict term of the sense, but also about the quality of anthropological and linguistic research. Although texts written by self-educated researchers such as Nimuendajú were still accepted at that time, Krause was a professional ethnologist and, moreover, a Prussian style public servant responsible for the utilization of the museum's annual budget. And he even was right with his doubt about the possibility of editorial interferences in Nimuendajú's texts, but not in the way he imagined. 12 Nordenskiöld replied on November 15 that, In the meantime, Nimuendajú and Krause had exchanged several letters trying to come to terms with Nimuendajú's detailed explanations about fieldwork possibilities and the interests of the museum. For example, Nimuendajú prepared a list of research possibilities in various regions of the Amazon and adjacent areas, in a letter dated August 12, 1927. 14 But the decisive factor was the museum's interest in acquiring ethnographic objects from the region between the Middle Tocantins River and the Mearim River, in Maranhão, with the aim "to fill regional gaps" in its collections. Moreover, Krause himself had carried out an expedition to the Karajá in the Araguaia region between January 1908 and February 1909. Thus, he saw in Nimuendajú's field activities a seamless continuation of his own research focus. 15 Unfortunately, not only Nimuendajú was short on cash, the Grassi Museum was also suffering from the permanent budget crises that hit most museums and scientific institutions in postwar Germany. In order to find a solution for the mutual interests, Krause succeeded, in longwinded negotiations, to convince Thilenius, from the Hamburg Museum of Ethnology, and Jacobi, from the Dresden Museum of Zoology and Ethnology, to support an expedition and share the costs. In return, Nimuendajú had to collect three copies of every object to be redistributed among the three museums. This obligation was even the subject 11 SES, Leipzig, 1928/43, Krause, November 12, 1927 12 In one of Nimuendajú's articles about mythology, religion and shamanism of the Xipaya Indians, the ecclesiastical editors of Anthropos did not revise any of his often popular expressions, but paradoxically interfered in the mythical text about the origin of the Brazilnut tree (Nimuendajú, 1919(Nimuendajú, /20: 1019, where one passage originally written in German was substituted by a translation into Latin. The reason? The original text contained some parts with explicit reference to human genitals. 13 SES, Leipzig, 1928/43, Krause, November 15, 1927 (translation P.S.; underline in the original letter) special issue | Peter Schröder | "Three Long Rows of Empty Shelves" to Fill: Curt Nimuendajú as Collector andResearcher for Ethnological Museums in Germany, 1928-1930 of a detailed contract, which Nimuendajú considered unnecessary because, from his point of view, he had to assume all the "real" risks, including his health and financial outcome.

the expeditions 16
On September 17, 1928, Nimuendajú wrote a letter to Krause explaining his apprehensions about the expedition's success: Finally, I have arrived at the conclusion, after having made purchases and having organized the exchange goods for the different tribes: Although I considerably exceeded the amount planned for this purpose, I became aware that it will not be sufficient for the complete execution of my program. I am probably not going to reach the Apinayé and the bands located to the northeast. I will have to give up going ahead to the Karaô and the Šerénte in the south, and the Canellas and Guajajára will probably come off somewhat scanty. -If it were only a question of simply illustrating the culture of the different tribes, everything would be very fine, but in my mind I see THREE long rows of empty shelves which I shall fill up all. Notwithstanding that, I believe optimistically that I will succeed to give satisfaction to all three institutes. 17 "In my mind I see THREE long rows of empty shelves which I shall fill up." This sentence seems to be a premonition of the difficulties Nimuendajú would face during the expeditions, even if collecting three copies of each object often was not the main challenge.
The first expedition started in September 1928 and finished in May 1929. Fieldwork was done among the Apinayé, Krĩkateyé, Kreapimkateyé, Pukobyê, Guajajara and Canela (Apanyekrã and Ramkokamekrã) in the current Tocantins and Maranhão states. Nimuendajú described the complicated fieldwork circumstances in his detailed letters to Krause, but also to Carlos Estevão (the latter case being in Portuguese). An example might illustrate the letters' style.
On October 31, Nimuendajú arrived at the Tocantins river: As can easily be suspected, the original aim of collecting three copies of each object was impossible to be achieved entirely, due to "real life" circumstances in the field, and had to be given up partially. The letters express profound cultural pessimism regarding the survival of the visited groups as culturally distinguished collectivities. This is explicit in the case of the Bacabal village of the Apinayé described by Nimuendajú in the same letter: The Apinayé gave me an extraordinarily friendly welcome. Their chief José Dias had prepared his people well over two years ago, when we had agreed upon a visit for the first time. It was as if I had returned to old acquaintances. Notwithstanding that, the Aldea [village] was for me, another one of the deceptions I am almost always used to have in such cases: Unfortunately, the old village with its bachelors' house and mask hut that the chieftain had drawn for me on a paper two years ago was no more. 5 huts in Neo-Brazilian style were arranged around a very badly cleaned, almost circular place; the population numbers less than 50 heads. Two years were enough to change many things: A wave of influenza had carried off the children and the young people. José Dias had undertaken a voyage to Goyaz and São Paulo in order to protest against the occupation of tribal lands by the neo-Brazilians. 19 He had already been declared dead and the Aldea had already disintegrated according to that rumor. The Brazilians just were about entering into his heritage when José Dias suddenly returned carrying a lot of gifts, which caused them a great displeasure. He joined his people again and build, not far away from the old one, the current Aldea da Bacaba [sic]; but the old Indian way is dead and gone. Fortunately, however, it is part of a past so recent that many things can still be recovered. José Dias as well as the old medicine man Nicolau are so friendly and eager that I believe that the voyage was worth the trouble. The circumstances are approximately the same as among Koppers' Yahgan 20 . Here I cannot walk from hut to hut and buy up a magnificent collection within 2 days, as you did among the Kayapó. I will have to dispense completely with triplicate collecting and would be very glad if I can at least succeed to more or less reasonably document the old culture. In about one month I will be able to inform you about the results among the Apinayé from the museum's point of view. There is no need to be too worried: It will not be very splendorous, but it won't be a fiasco. The situation now calls for waiting patiently until the old customs and objects occasionally turn up again. The chief insists on that I should stay AT LEAST for a month. Then I would still get to see many interesting ceremonies. It would not be possible without further ado. It stands to reason that we also cannot celebrate the Midsummer Day and Christmas within a week. The old medicine man asked intently if I also had brought along with me enough paper: I ought to have such a high pile so that he could dictate me EVERYTHING; now he also wants to give me his younger daughter in marriage.
19 Neo-Brazilians is a word used by Nimuendajú which refers to the non-indigenous population in Brazil. Implicitly, the indigenous peoples were 'Brazilians' or, with other words, the 'real' Brazilians. This kind of classificatory scheme (Brazilians -neo-Brazilians) has its origins in Nimuendajú's version of Americanist Ethnology with its suppositions about the difference between 'original' and 'mixed' , 'acculturated' or even 'deformed' cultures. Some traces of German romanticism may even be identified in such a classification. By the way, Nimuendajú's scheme neither became established in Brazilian anthropology nor by South Americanist ethnology. It remained part of a personal vocabulary.
20 Reference to Koppers (1924) and his ethnography about indigenous peoples in Tierra del Fuego. special issue | Peter Schröder | "Three Long Rows of Empty Shelves" to Fill: Curt Nimuendajú as Collector and Researcher for Ethnological Museums in Germany, 1928Germany, -1930 Rev. antropol. (São Paulo, Online) | v. 62 n. 1: 217-240 | USP, 2019 Nimuendajú patiently waited until old objects re-emerged, but he also induced the Apinayé to fabricate them again. He tried, for example, to resuscitate the manufacture of masks, but was unsuccessful. In the end, he left the village with some 300 objects. However, he became more enthusiastic after having initiated fieldwork among the Ramkokamekrã whose culture he considered a kind of "gold mine" for ethnographic research:

Ponto, the village of the Ręmkókamekra is the biggest Indian settlement I have ever seen: it comprises 31 huts with more than 300 residents. […] In a word, the Aldea do Ponto is an ethnological treasure: it is also the only Aldea where one nowadays can still study in their entirety the social and religious institutions of the Timbira. 21
In his eyes, however, this Canela group was constantly threatened by non-indigenous influences of all kinds or, in his own words, by "the Christians," "Christianity" or "the neo-Brazilians":

Two circumstances render it particularly difficult to take a longer sojourn in this village, which is so interesting: In the first place, the brandy dealers who plague the Aldea every other day; in the second place, the excessive begging already mentioned. Indeed, it is not easy to live for a month among a bunch of 300 beggars and to have to keep them in a good mood! [...]
By the way, once again I have been very successful in achieving one thing: The friendship of the Indians of Ponto; and in the same measure as it increased, the begging diminished somewhat. They almost fed me to death; on the central ground of the village they solemnly gave me the name of their deceased principal chief Kukaipó; unlike the case of other tribes, I was the declared favorite of the women and girls who decked me out with adornments like a Christmas-tree and devoted hours to painting me very properly from top to toe; as luck would have it, they did not give me in marriage as the Apinayé. Again and again they said that I should stay, that I should attend the youth initiation ceremony, enter in an age class -Lord, and protect me from my friends! Because, after having first kicked out 3 brandy dealers one after another as if I was entitled to do that, the Christianity on all sides got scared and did not dare any more to appear with brandy.
Sometimes his letters also include comical observations, for example about the Guajajara, neighbors of the Canela: "Bananal [village] was still more civilized than Lagoa da Pedra. There, I also had the pleasure, among other things, to see Indian wives with bobbed hair (Bubiköpfe) dancing tango while being accompa-special issue | Peter Schröder | "Three Long Rows of Empty Shelves" to Fill: Curt Nimuendajú as Collector andResearcher for Ethnological Museums in Germany, 1928-1930 Rev. antropol. (São Paulo, Online) | v. 62 n. 1: 217-240 | USP, 2019 nied by accordion, and they pleasantly invited me!" 22 On the whole, the expedition was highly successful from the point of view of the three museums, but the collector, in his own descriptions, hardly managed to return to Belém because of the hardships in various field situations. He returned home quite ill and without a penny in his pocket, so that the three museums decided to arrange some extra resources in an emergency action between the 5 th and 12 th of June, 1929.
One month after Nimuendajú had concluded the expedition, Krause asked in a letter from June 27, 1929, about his future plans and started to suggest another contractual work for German museums and research institutions. 23 In a letter from May 31, 1929, Krause had already set out his plan for "the further fate of Mr. Nimuendajú" to Arnold Jacobi from the Dresden Museum of Zoology and Ethnology. 24 As the museum's resources would be insufficient, he suggested to involve the Saxony State Research Institute for Ethnology (Staatlich-Sächsisches Forschungsinstitut für Völkerkunde, 1914-1936 of Leipzig University. Ultimately, the second expedition was financed by the ethnological museums of Leipzig and Hamburg, by the Institute of Ethnology of Leipzig University and with the support of the Emergency Association of German Science (Notgemeinschaft der Deutschen Wissenschaft, NDW), founded in 1920. 25 During this expedition, which lasted from February to September 1930, Nimuendajú carried out fieldwork among the Apinayé, Xerente, Krahô and, again, the Ramkokamekrã.
His descriptions of indigenous peoples present a pessimistic view. During the first expedition ethnographic information was only requested for completing the description of the objects collected; during the second expedition the writing of a monograph (with the title Die Timbira) was agreed upon. This should be a complementary publication for the planned exhibition of the objects. This monograph, about the Apinayé and the Canela, was the subject of protracted negotiations between Krause and Nimuendajú, but ultimately was not published as originally planned. The financial restraints of German scientific institutions became so critical in the final years of the Weimar Republic that this plan had to be abandoned. Nimuendajú became thoroughly dissatisfied and this led him, among other things, to gradually withdraw from German ethnology and also prepared the ground for his contacts with American anthropologists. Die Timbira was finally published in 1939 and in 1946, but not in its original form. The Apinaye and The Eastern Timbira, two ethnographic studies translated and revised by Robert Lowie, originally planned as two parts of the monograph, are nowadays deposited in two copies (one at the Museum of Ethnology in Dresden, the other at the National Museum in Rio de Janeiro). 26 It is interesting to note, though not surprising, how the indigenous peoples visited during the expeditions were seen by the parties of the contracts. Already, 22 SES, Leipzig, 1928/43, Krause, February 15, 1929 23 SES, Leipzig, 1929/71, Krause, S. 700, June 27, 1929. 24 SES, Leipzig, 1928/39, Krause, May 31, 1929 25 The Emergency Association of German Science (NDW) was founded on October 30, 1920 as a centralized institution with the objective to facilitate raising and distributing of resources for German scientific research. Due to the general financial situation after World War I, NDW had an important role in the general context of decay scientific research was suffering during that period. In 1933, NDW was taken over by the Nazis and ended its activities in 1945. The current DFG (Deutsche Forschungsgemeinschaft/ German Research Foundation) was founded in 1951 in the Federal Republic of Germany as the successor of NDW. special issue | Peter Schröder | "Three Long Rows of Empty Shelves" to Fill: Curt Nimuendajú as Collector andResearcher for Ethnological Museums in Germany, 1928-1930 a superficial reading of the correspondence permits the conclusion that the German museum staff, including Krause with his own field experience, saw the Amerindians primarily as suppliers of interesting objects for the museum's depots. Nimuendajú, on his part, once having entered "the field," adopted a very different position, assuming a role that nowadays would be called engaged anthropology.
As in the case of other fieldwork he conducted, Nimuendajú generally played down the results, transforming them into "failures" or at least "small successes" before receiving any written reaction of the addressees of his letters and shipments (with the objects). This seems to have been a kind of preventive strategy, as he was well aware of the high expectations of European museums, where armchair anthropologists often did not have an idea of the difficulties that arose in the field (even if this does not apply to Krause).

implicit theories and explicit methodology
Is the history of the two expeditions only an interesting example of anthropological collecting in the twentieth century, or were there any theoretical and methodological aspects that could be taken from them?
Nimuendajú himself did not have theoretical pretentions, as was noted by several authors (Welper, 2002). His texts are frequently characterized as purely descriptive, and Nimuendajú himself contributed to this image of an ethnographer uninterested in theoretical questions. Indeed, he never wrote any theoretical text, which can be explained by his lack of formal academic training, and consequently by his fear to expose himself unnecessarily in academic circles.
Yet, organizing collections for museums does not happen in a theoretical or methodological vacuum. As in the case of ethnographic studies, a pure description does not exist (neither of objects nor of cultural practices); any description is based on theoretical choices, at least implicitly (Bruck, 1987). Composing museum collections and selecting its objects implies taking decisions based on explicit and implicit theories (Clifford, 1988). The history of Nimuendajú's collections in the Hamburg, Leipzig and Dresden museums confirm this affirmation. In this case, we have to take into account his acquaintance with contemporary German ethnology; that is, his lectures and correspondence with German ethnologists at the time.
In his published texts, as well as in his correspondence we can find references to ethnological historicism, with its methodological principle of a particularizing "research of facts" (Müller, 1980). Alleged theoretical abstinence even formed part of recurrent recitations done by many German and Austrian diffusionists as a discursive effort to banish the vices of speculative classical evolu-the museum in September, 2018. Maybe the existence of the copies nowadays might excite the ambition to publish the monograph The Timbira in Portuguese, German or English or in a bilingual edition. However, it should be taken into consideration that the mere publication of the manuscript, though tempting in times of neoliberal productiveness, does not make any sense without a careful comparison with the subsequent published monographs of 1939 and 1946. And this is only possible with intimate knowledge of the German language. special issue | Peter Schröder | "Three Long Rows of Empty Shelves" to Fill: Curt Nimuendajú as Collector andResearcher for Ethnological Museums in Germany, 1928-1930 tionism. can The present author still observes this discursive practice among the last representatives of German diffusionism during the 1980s.
Although organizing collections for European and Brazilian museums was not Nimuendajú's only activities during the 1920s and 1930s, collecting objects was an essential part of his ethnological practice, and this can only be understood in the historical and theoretical context of contemporary German ethnology. Purchasing ethnographic collections through research expeditions was not merely part of the rivalry between ethnological museums in imperial and post-imperial Germany (Penny, 2002). In addition, research following Adolf Bastian's ideas about collecting ethnographic artifacts included the thematic priority of material culture, as was demonstrated by Michael Kraus (2004) in his brilliant study of German ethnological research in the Amazon region between 1884 and 1929.
Krause's generation regarded the collection of objects as inseparable from the concept of doing ethnography, because the objects were interpreted as sources containing ethnographic information. By that time, Brazil held an important position among the purveyors of objects for foreign museums. Castro Faria (1987) even recommended that Nordenskiöld was a protagonist in the mercantilism of archaeological and ethnographic collections, even if this practice was indeed much older. Nimuendajú's collecting, however, was not merely commercial, as his collections became particularly valuable because of the detailed ethnographic information supplied with the objects. His ethnographic practice had much more to do with contemporary ideals of cultural salvage in German ethnology than with predominant theoretical concerns. From his point of view, it was urgent to record indigenous cultural manifestations before they were abandoned and forgotten. Thus, we can recognize in Nimuendajú a greater affinity with ideas of Adolf Bastian and Franz Boas than the other directions in the anthropology of his time (Fischer et al., 2007;Stocking Jr., 1996).
Other indications of theoretical interest in the history of the two expeditions are Krause's brief comments about the ethnographic collections at Leipzig and Dresden, which would need to be "completed" for two main reasons: (a) the lack of "material" from whole "cultural areas or provinces" (a diffusionist argument) and (b) the idea that cultures can be represented by a sensible and discerning selection of objects for offering a "complete" image of an indigenous culture by its materialized aspects. The second point of view was completely shared by Nimuendajú.
While anthropological theories remained implicit in the correspondence about the expeditions, methodological aspects became quite explicit, above all when one takes into account contemporary standards for reporting fieldwork activities. Nimuendajú's letters from the field are quite detailed, very personal, and direct about the situations in the field and their vicissitudes. We do not find any objective, scientifically weighed discourse, but rather vivid special issue | Peter Schröder | "Three Long Rows of Empty Shelves" to Fill: Curt Nimuendajú as Collector andResearcher for Ethnological Museums in Germany, 1928-1930 descriptions about the difficulties and hardships experienced. Father Wilhelm Koppers (1886Koppers ( -1961 from the journal Anthropos even decided to publish a private letter sent to him by Nimuendajú from the field about his difficulties in meeting the Canela's constant requests for presents (Nimuendajú, 1929b), but this went too far. After a protest formulated by Krause 27 in the name of the three museums, Koppers felt obliged to publish a formal explanation about his reasons for having published Nimuendajú's letter in Anthropos (Nimuendajú, 1929(Nimuendajú, , vol. 24: 1104. Krause had informed him that the directors of the three museums were anxious that Nimuendajú's letter could give the impression they were exploiting the willingness of a cooperative researcher with insufficient means, leaving him destitute and ill after having received rich collections organized in very adverse circumstances; whereas Koppers firmly denied that such a negative interpretation could be imperative. While the three museums and the journal came to terms with the publication of a formal explanation, Nimuendajú never forgave Koppers, seeing his decision as an indiscretion. He broke off contacts with the periodical and never again published a text in Anthropos. Nimuendajú's fieldwork style can be described as radically individualistic, solitary and highly empathetic. In his extensive, but often prolix thesis about Nimuendajú's fieldwork methodology, Dungs (1991) pointed out that Nimuendajú did not only use to "live among the Indians," but that he preferred to live "like an Indian." Welper (2002: 127) emphasized his conviction that Nimuendajú needed a kind of cultural and social conversion as a precondition for successful fieldwork. Indeed, "living among the Indians like an Indian" implied much more than the scientific collection of information. It established social relations and implied obligations and moral engagements not intended in contemporary idealized views on scientific work. Thus, Nimuendajú's ideal fieldwork strategy can best be understood as being a mix of a lone wolf (Einzelgänger, in German) and a border crosser (Grenzgänger) (Figure 4). 27 SES, Leipzig, 1928/43, Krause, October 19, 1929partially reproduced in Anthropos, 1929, p. 1104. special issue | Peter Schröder | "Three Long Rows of Empty Shelves" to Fill: Curt Nimuendajú as Collector andResearcher for Ethnological Museums in Germany, 1928-1930 Rev. antropol. (São Paulo, Online) | v. 62 n. 1: 217-240 | USP, 2019

the collections today
The original lists of objects archived at the Grassi Museum in Leipzig indicate 2,478 ethnographic and archeological objects collected by Nimuendajú. In addition to objects from the Apinayé, Canela, Guajajara, Krahô, Kreapimkateyé, Krĩkateyé, Pukobyê and Xerente, there were also a large number of archaeological items. In addition, Nimuendajú offered the first collection of objects to the museum in November 1927. Although not too much is known about that earlier collection, the number of 2,478 objects deposited remained constant until December 4, 1943, when Leipzig suffered a heavy bombing raid and one bomb hit a wing of the museum where numerous objects had been exhibited. Only 613 objects remained undamaged. In other words, the Nimuendajú collections in Leipzig suffered a 75% loss in just one night of World War II. Thus, the collections at the Grassi Museum were the most affected by war among the three museums ( Figure 5).
The Dresden museum, on its part, received 354 objects (Apinayé, Canela-Ramkokamekrã, Guajajara, Kreapimkateyé, Krĩkateyé, and Pukobyê) from the first expedition. Although that collection was not so tragically affected as the one in Leipzig, the total losses were 81 objects, or 23% (KV = Kriegsverluste = "lost by war": 81; other objects are registered as "missing" / fehlt: 5). Currently, the Nimuendajú collection at Dresden comprises 273 objects, and some of them were exhibited during an exposition about Amazonian indigenous cultures in 2009-1010(Kästner, 2009.
According to an Excel list received by e-mail on October 13, 2010, the Hamburg Museum of Ethnology must have received three contingents of objects on three different data. According to the museum's books of register these include: By studying these registers and comparing them to another Excel file from Hamburg received in 2011, it becomes clear that the Nimuendajú collections in Hamburg represent the most complete and best preserved of all three col-

Figure 4
The Grassi Museum, nowadays (Courtesy: GRASSI Museum für Völkerkunde zu Leipzig, Staatliche Kunstsammlungen Dresden; photo: Volkmar Henze) special issue | Peter Schröder | "Three Long Rows of Empty Shelves" to Fill: Curt Nimuendajú as Collector and Researcher for Ethnological Museums in Germany, 1928-1930 lections resulting from the two expeditions. However, direct access to these collections in Leipzig, Dresden, and Hamburg, may be cumbersome and require a great deal of time for researchers because of the bureaucratic procedures that can be quite extensive.
Yet, Nimuendajú assembled two other collections for German ethnological museums, namely for those in Berlin and Munich. The collection in the Ethnological Museum of Berlin, organized in 1935, originally comprised 228 objects from the Ramkokamekrã detailed in the entry logs and the complementary documentation. However, during a visit to this museum in February 2011, only 135 objects could be located in the cabinets of the enormous depository, where approximately 35,000 objects from the South American lowlands are now preserved. The history of the Berlin collection was the subject of two papers presented during two congresses (Schröder, 2013;2017) and will be published in a forthcoming article.
By contrast, Nimuendajú did not directly sell the collection of 86 objects from the Ramkokamekrã at the Ethnological Museum of Munich to the museum; but, in August 1933, he had offered it to a German aviator called Otto Meyer who was living in Pernambuco. When Otto Meyer returned to Germany in 1934, he first offered the entire collection to the museum in Munich, but later decided to sell it to Gustav Umlauff, the well-known ethnographic object-trading firm in Hamburg. This firm then resold the collection, with the exception of some objects, to the Ethnological Museum of Munich in October, 1934. Single objects collected by Nimuendajú that can now be found in other German museums (that of Hannover, for example) were generally acquired by exchange between German and foreign museums (the Ethnographic Museum of Gothenburg, for example).
The histories of the Berlin and Munich collections and of the single objects collected by Nimuendajú available in other German museums are not the subject of the present article, as they resulted from a post-1930 period in Nimuendajú's collecting activities and were not part of the contractual agreements of the two expeditions he carried out in 1928-1930. conclusion What kind of anthropology was practiced during the two expeditions of 1928-29 and 1930? Was it German anthropology at the southern fringes of Amazonia, the initial stage of Brazilian anthropology, or something else? This depends on the point of view taken. In Brazilian anthropology, Nimuendajú has long been incorporated as a prominent person in its genealogical tree. From this perspective, the question would be easily answered, although it is difficult to neatly differentiate a national anthropological tradition in Brazil for the period around 1930. At that time, curious individuals generally practiced Brazilian anthropology with academic backgrounds, lacking formal anthropological training, and there was no intellectual school or tradition that could be followed. Scholars such as Edgar Roquette -Pinto (1884-1954 or Capistrano de Abreu (1853-1927 were exceptions to the rule. As a self-educated researcher, and an enthusiastic and studious reader of ethnographic studies published by German ethnologists, Nimuendajú repeatedly deplored in his letters that sometimes he had only access to popular (reduced and simplified) versions of German publications about the indigenous cultures of the South American lowlands. Notwithstanding, German ethnology was his main anthropological reference at the time. Thus, around 1930, he was above all a German expatriate living in Brazil, practicing Americanist ethnology without being part of a national academic establishment. Gradually, however, he became part of an international transatlantic network of knowledge interchanged by academic and non-academic anthropologists, archaeologists, and linguists.
And what might have been his own answer? Nimuendajú's correspondence with Koch-Grünberg makes this quite plain: The "we" used in many letters refers not only to a common nationality, but also to the mutual comprehension of a national anthropological tradition. However, the decline of German ethnology's role in Brazil from the 1930s onwards, coincides with Nimuendajú's expeditions. Therefore, the lack of possibilities to continue researching and collecting ethnographic objects for German institutions prepared the ground for an intensive and productive cooperation with American anthropologists, especially with Robert Harry Lowie, and also for systematic research about Jê-speaking indigenous peoples. It is interesting to observe that Nimuendajú received suggestions to read some of Lowie's works in the correspondence with Krause (Lowie, 1920;1924), 28 which were even ordered for him in London and were included in the budget for the second expedition. Herewith, we can observe a parallelism between a change in Nimuendajú's scientific biography and a macro tendency in the anthropology in Brazil during this period. That is seen as a gradual distance from German ethnology with a simultaneous approximation to North-American anthropology.
As in the case of other ethnographic collections, the history of Nimuendajú's two expeditions financed by German institutions reveals that these collections were more than a set of objects whose organization was based upon some explicit or implicit criteria of the collectors and their financiers. With their objects and their complementary documentation, they offer differentiated information and insights about various aspects of anthropological practices in specific historical contexts. Therefore, the collections and their documentation become sources that "speak" as much about the indigenous producers and users of these objects, as about the anthropologists involved in collecting them.