Although the roots of geography, as a field of study, reach back to Classical Antiquity, (See Supplementary Note 2) its establishment as a modern science was essentially the work of the century from 1750 to 1850. The second half of this period, the time of Humboldt and Ritter, is commonly spoken of as the "classical period" of geography. Undoubtedly the extraordinary accomplishment of each of these men, working at the same time but in very different ways, and the influence of their work on all subsequent geography justifies our regard-ing them as the first masters of modern geography--in that sense as the "founders." But in applying such titles one is easily led to ignore the importance of previous workers--those, we might say, who laid the foundations for the founders. "Carl Ritter hat auch seine Vorganger," Peschel is quoted as saying, and the statement applies as well to Humboldt. Among these predecessors we are not here concerned with the geographers of Classical Antiquity nor with the majority of the writers of geographical works during the two or three centuries preceding 1800. Our interest is with that smaller, but still considerable, number of students of the second half of the eighteenth century who consciously strove to convert a more or less miscellaneous and useful study into an independent science. Although one could hardly affirm that they had accomplished that purpose, nevertheless it is clear that they had put down the main outlines of the science of geography, as we know it to-day, before the appearance of geographical publications by Ritter and Humboldt at the turn of the century. They had not established those outlines, however. It is for that work, the establishment of the modern science of geography, that we are indebted to Ritter and Humboldt. Nevertheless, most of the fundamental concepts of geography--including almost all of those which we will have to examine in this paper--may be found in the writings of the German geographers immediately preceding Ritter and Humboldt. It hardly seems too much to say that, had neither of these men lived, the development of geography after 1800 would have led ultimately, even though far more slowly, to something like that which we know.

As the inheritors of the geography of Classical Antiquity and that of the Renaissance, the geographers of the eighteenth century as a matter of course considered all kinds of phenomena whose differences in different parts of the world made them appear significant to a knowledge of the world. They echoed the famous statement of Strabo: "Geography, in addition to its vast importance to social life and the art of government...acquaints us with the occupants of the land and ocean and the vegetation, fruits, and peculiarities of the various quarters of the earth, a knowledge of which marks him who cultivates it as a man earnest in the great problem of life and happiness" [38, introduction].² The "Physical Geography" of Immanuel Kant was concerned with all the elements included by Strabo; on the basis of this physical geography, Kant would found "other geographies" including political, commercial, moral (in the sense of mores), and theological geography, as well as the descriptions of each particular land [40, § 5]. Physical geography is the essential propaedeutic for an understanding of our perceptions of the world, whether those received directly by travel or indirectly by reading [§ 2, end of § 4]. In brief, "it serves as a suitable arrangement for our perceptions, contributes to our intellectual pleasure, and provides rich material for social discourse" [§ 5; though the ideas are presumably those of Kant, the words are in the voice of Rink, see footnote 3].

These statements of Strabo and Kant are susceptible of translation in terms of modern science. What Leighly refers to as "the unspecialized curiosity in the minds of stay-at home readers" was not considered by Kant as in opposition to science. On the contrary the curiosity which desires a knowledge of the world is the stimulus that requires science to pursue that knowledge and seek to organize it. For this purpose, to be sure, it is necessary that individual students and groups of students should specialize, but Kant recognized that there is more than one form of specialization. While he recognized the importance of the specialization according to kinds of objects (which was later to dominate the field of the natural sciences), he saw that that form of specialization was not possible in geography, since its single object, the earth, is for us unique in the universe. We shall consider later the form of specialization which he outlined for geography (Sec. IV B).

The great majority of the writers of geographical works of the seventeenth and eighteenth centuries, however, did not attempt to translate the purpose of geography in terms of scientific interest. Rather, they were concerned with its practical utility. Over and over again they emphasized the value of the study of geography as a means for other purposes--for an understanding of history and as a practical aid to government. Wisotzki has shown how this utilitarian point of view prevented scientific progress in geography [1, 96-130].

Overlooking earlier but unsuccessful efforts, we need to consider a small but increasing number of writers in the second half of the eighteenth century who worked to elevate geography from its subordinate position to that of an independent science, "reine Geographie," or "pure geography," as they came to call it. (In contrast with later uses of this term, both in Europe and America, the reader will note the suitability of its use in this connection.)

This movement found its first expression in the demand that geography should consider its object, the world, in terms not of political divisions but of more real and lasting divisions of nature. Divisions of area can be established only by boundaries, so that this concept found its first expression in a demand that geography should divide the world by "natural boundaries" of "lands" rather than by political boundaries of states. This movement, which represented an echo from Strabo, was expressed by many writers, both geographers and students of law (notably Grotius), as early as the 16th century [1, 193 ff.; 5, 44 ff.]. It could make little headway, however, so long as geography was considered to be merely the handmaid of history and government. Although it was argued that the frequent changes in political boundaries made the geographer's work seem of but transient value, the argument was irrelevant so long as it was the major business of the geography teacher to provide information about political areas. To accomplish this task at a time when the political map of central Europe--Germany and Italy--was a crazy-quilt of hundreds of states, the geography teacher could hardly spare time to consider "natural regions."

On the other hand it seemed obvious to those who were attempting to develop a scientific geography, that the conventional organization of the knowledge of different parts of the world by political divisions was anything but satisfactory. Whether in terms of the political map of Europe of that day or this, or of the political map of the United States today, it hardly seems too much to say that no student who is seriously concerned with the attempt to develop regional geography could regard political divisions as even "generally satisfactory."

Then as now, however, the effort to consider the areas of the world in divisions other than political presented the difficult problem of finding an equally definite method of dividing areas of the lands. The answer to this long-felt need appeared to be offered by a theory put forward by various students in the middle of the eighteenth century, notably Buache in France--namely, the theory of the continuous network of mountain systems [see Wisotzki's chapter on "Der Zusammenhang der Gebirge," 1]. The theory was immediately taken up in Germany, and Gatterer made it the basis of a physical division of the world into lands and regions (Gebiete). Whereas the similar attempts made much earlier--in Holland and Italy as well as in Germany--had been largely ignored, Gatterer's work had a marked influence on his contemporaries and on later students, including Humboldt [cf. Richthofen, 3, 671). We may therefore mark the beginning of a continuous reformation of geography with the appearance of his publications in 1773-75 [1, 201 ff.].

A somewhat different stimulus to the development of scientific geography throughout this period came from the lecture course on "physical geography" that the philosopher Immanuel Kant presented at the University of Konigsberg, repeated some forty-eight semesters during the period 1756 to 1796 [14, 9 f.]. For Kant himself, as Gerland has noted, the study of geography represented only an approach to empirical knowledge necessary for his philosophical considerations [12]. But finding the subject inadequately developed and organized, he devoted a great deal of attention to the assembly and organization of materials from a wide variety of sources, and also to the consideration of a number of specific problems--for example, the deflection of wind direction resulting from the earth's rotation [39; 13].

It is difficult to estimate the historical importance of Kant's work in our field. His course in physical geography, one of the most popular of his courses, was heard by a very large number of students during its forty-year period, and long before the publication of the lectures, after 1800, there were many fairly complete handwritten copies in circulation. Furthermore, former students, like Herder, had given wider circulation to his ideas. Whether by indirect means, or only after their publication in 1801-05, the lectures had a significant influence, as we shall see, on both Humboldt and Ritter.

The greater part of Kant's course followed the outline of "general geography" that Varenius had laid down a century before and which Lulofs had followed much later (in German translation, 1755). In addition to the large amount of material taken from Lulofs, Kant later drew from the work of Bergman (published in Swedish, 1766, in German, 1769), as well as Buache, Buffon, and especially Busching [40, notes, 551 ff.; 14, 285 ff.]. In his relatively briefer consideration of individual countries, however, he does not appear to have been stimulated either by Buache or Buffon to have altered Busching's conventional division by political units; certainly the outline and material of the course were well established before the appearance of Gatterer's work [39, 11, 1-12; 14, 27, 31 f.; 15).³

We cannot here consider the various steps in the subsequent development of the "pre-classical" period, and can only briefly suggest the underlying factors that stimulated its progress. Of great importance was the rapid growth in various physical sciences. In particular the development of reasonably accurate barometric methods, notably by de Luc, whose work was first published in 1772, made it possible to measure the "vertical" dimension of the earth's surface [Marthe, 25, 30; Peschel, 66, I, 329 f.]. The fundamental approach of many geographers of this and the following period was notably influenced by the great reformer of education of the time, Pestalozzi, of Zurich, who brought into German schools the philosophical viewpoint of Rousseau [25, 30; 1, 260 ff.]. Closely related to the change in geographic thought--whether as cause or effect--was the so-called "natural division" of France into departments according to rivers and mountains. Somewhat later the radical and repeated changes in political boundaries during the Napoleonic Wars helped to discredit geography texts based on such fluctuating conditions, and thus helped to make way for a geography based on the physical description of the world [1, 257-66].

Our present interest is in the nature of the field of geography as it was conceived by a small but active minority of writers in the late eighteenth century who, before Humboldt and Ritter, were endeavoring to establish geography as a science. We need not attempt to set a definite limit as the end of this period, but may consider it as overlapping perhaps a decade or so into the nineteenth century. Although Humboldt published at least one geographical article before his all-important journey to tropical America in 1799, his major publications in our field did not begin until after his return in 1804, the year in which Ritter's name first appears in geographical literature. According to Marthe, however, it was not until nearly 1820 that each of these two figures assumed a dominant position in geography [25, 6].

The geography which the late eighteenth century had inherited from the past was essentially limited to phenomena of the world as known to man, that is, of the earth's surface, in the broad sense of a surface of some thickness extending into the atmosphere and into the solid earth wherever man had penetrated. Kant had described the field of physical geography as the world (Welt), so far as we can come in relation to it, the scene of our experiences. The study of the earth body belongs to mathematical geography, but is considered in physical geography insofar as it causes differences in its different parts [40, 1, §2, 7). In most of the geographies of this period the treatment of the earth-body was confined to an introductory chapter--"Mathematische Vorbegriffe," according to Kant--and the actual study of geography confined to the earth-surface. The specific term "Erdoberfläche" was frequently used, and in 1820 Wilhelmi specifically limited the field of geography to the earth-surface.⁴

The geography of this period included both "general" studies of particular kinds of phenomena of the earth surface, and descriptions of many kinds of phenomena found in particular areas. The inclusion of these two forms of study within the single field has no doubt been the cause of more controversy than any other single problem in the methodology of geography. It is significant therefore that it was not introduced into modern geography as a result of the chance combination of Humboldt and Ritter. On the contrary the same difference is found in the work of the geographers of antiquity, as Hettner observed in his first brief treatment of the history of geography [2, 306 f.; more fully, 161, 33 f.]. The tendency of these two directions to come in conflict with each other and to interchange positions within geography at different times is evidence, he suggests, that they do not represent separate sciences but merely different directions within the same science. The distinction between the two directions was perhaps first clearly stated by Varenius (a German, Bernard Varen, living in Amsterdam) in 1650 [10, 100 ff.]. In contrast with most of the writers of his time who limited geography largely to a bare description of the several countries-including therein much of their political constitutions--Varenius divided the field into "general or universal geography" and "special geography" or "chorography" [note Humboldt's discussion of Varenius, 60, I, 60]. His untimely death, at the age of twenty-eight, in 1650 prevented him from studying the second part, though his previous work on the geography and. history of Japan would indicate that he did not lack interest in that part of the field.

The terms which Varenius had used, "general" and "special" geography, later became the standard terms in Europe for these two aspects of the field, though many later writers have found them unsatisfactory. As we will note in a later section, the frequent use by German writers of the adjective "systematisch" in describing "general geography" supports the common use in this country of the term "systematic geography." The term "special geography" was largely replaced in German literature by "Landerkunde," which in spite of obvious disadvantages is commonly favored over non-Germanic terms, "special," or--the term now nearly universal outside of Germany--"regional geography" (see Sec. XI E).

Following Varenius, Lulofs and Bergman had concentrated on systematic geography. Kant, however, though depending in part on them, likewise suggested the place of Länderkunde, but his beginning of such a regional study is not firmly based on his systematic studies [40, I, §3, II, Abs. 3]. The distinction between the two aspects of geography was more explicitly stated by Gatterer, 1773-75, by Krug, 1800, and particularly by Bucher, 1812.

Although the systematic studies of this period undoubtedly were of major importance in the progress of geography toward scientific standing, Plewe finds that many of them, notably those of Bergman, tended to fall out of geography and become studies in other sciences. In contrast, John Reinhold Forster--the first of the great scientific geographic travelers--consistently maintained the "macroscopic" sense suitable to geography and endeavored to explain the regional relations of each type of phenomena studied. The work of J. R. Forster, and particularly that of his son, George, was regarded as of great importance by Humboldt, upon whom, as we shall see, they had a determining effect. Plewe, among others, counts the elder Forster as the first --in order of time--of the great "classical geographers" [8, 22-26].⁵

The division between general or systematic geography and Länderkunde, or regional geography, represented therefore a form of dualism that was characteristic of geography throughout its initial period of development as a modern science. This form of dualism in geography is not to be confused, as is often the case, with a dualism in terms of content, as Wagner warned, in 1890 [80, 375].

In terms of content, the eighteenth century writers commonly distinguished three divisions: "mathematical geography," "physical geography," and what was variously called "historical" or "political geography," but which is not to be confused with the meanings of those terms as used today (or as used by Kant). Mathematical geography consisted in large part of the study of the earth as an astronomical body. As we have seen, it was always of minor importance and in practice its study was left largely to astronomers. Zeune, in 1808, and Butte, in 1811, definitely excluded it from their outlines of geography.

The major contrast in content was that between what was included in physical geography and that contained in what we might call social geography. With the much more rapid increase in development of scientific methods of studying physical, as distinct from social, phenomena the contrast between the two became more marked. Attempts to develop a unified science included suggestions of confining the field to what was then called "physical geography." But if the present-day student considers this period only in terms of the titles which its writers used, he will have an erroneous conception of their works. As used by nearly all the writers of this period--including Kant, Forster, and later, Humboldt--the term "physical geography" was not limited to that which later was to be called the "physical" or "natural" environment, but included races of men and, commonly, their physical works on the earth. Thus Kant not only included man as one of the features "encompassed in the earth surface (Erdboden)" but also considered man as one of the five principal agents affecting changes on the earth [40, II, §1-7, I, §74]. Indeed, Kant's "physical geography," both in purpose and in content, might be considered as "anthropocentric," a point of view which Ritter inherited from Kant [according to Becker, 5, 53].

Very few writers of this period made the distinction which is most familiar to us, between a geography of natural, or non-human features, and a human geography, though it may be found in the work of J. M. F. Schulze, 1787, and of Rühle, 1811. For most writers of this and the following period--i.e, until perhaps the middle of the nineteenth century--the term "nature" (Natur) was not used in contradistinction to human, but to indicate that which was perceived externally in contrast to one's internal feelings and thoughts. So far as I can find, no geographer of this period questioned the place of either human or organic features of the earth in the field of geography. One might note, however, a trend in that direction, in the interpretation which Hommeyer, in 1810--and perhaps others--gave to "reine Geographie," namely, as limited to the conditions of the terrain; the study of climates, minerals, and organic life was left to "Naturkunde," "Naturbeschreibung" and "Länderkunde" [1, 221].

Although the writers of this period recognized that the scientific geography which they were attempting to develop included somewhat separate parts--"physical" and social geography--they considered these to be associated not merely because the phenomena were to be found in the same places, but because they were causally connected by mutual relationships. No doubt the interest in relationships between different groups of men and the particular character of their natural environment has ever been considered appropriate in geography. Kant, for example, noted that "in the mountains, men are actively and continuously bold lovers of freedom and their homeland" [40, II, §4], and similar relations were discussed by Gatterer and his followers [5, 48-53). Kant specifically based the social branches of geography on their relations to physical geography: "theological principles in many cases undergo important changes as a result of differences in the land (Boden) " [40, I, 19). In this he was followed by Müller, in 1785, and Fr. Schultz, in 1803, and by many others in the following decade. So far as I can find, however, it did not occur to any of the students of this period to consider these relationships as in themselves the direct object of geography study.

During the latter part of this period, it became common to regard the intricate interlacing of relationships between all the phenomena found in the earth surface, organic and inorganic, as functional relations of parts of a single whole. The concept was early expressed by Kant, for whom the systematic arrangement of objects by classes, as in the system of Linnaeus, divided nature into parts without organizing a system,⁶whereas physical geography "gives an idea of the whole, in terms of area" (Raum) [cf. Plewe, 8, 39]. At the beginning of the following century this concept was apparently so strongly supported by the philosophical views of the day that Butte stated, in 1811: "no scientist doubts the reality of an earth organism" [1, 230]. Similarly, within any particular area, the combination of all interrelated phenomena is not a mere aggregate but an interrelated "Whole," according to Krause as well as Butte, in 1811.

Few of the writers of that period--or indeed of any later period--distinguish clearly between the concept of unity of all the phenomena at any particular place or area, in what we may call a vertical totality or unity, and the horizontal unity of the area as an individual unit distinct from neighboring units. This latter concept was, as Bucher noted in 1827, in a sense forced upon the promoters of the new school of geography in competition with the old [51, 86 f.]. In place of the definite areal units of states, sharply defined by political boundaries, the new geography required equally definite "natural" units, somehow defined in nature. For a time, such definite natural boundaries appeared to be provided by drainage basins sharply separated by the "network of mountains," and, of course, by the seas. As increased knowledge of the actual conditions of the earth's surface made this theory untenable, the problem of finding "natural boundaries" for such "natural units" of area became much more difficult. We need not here concern ourselves with the long conflict between the supporters of "dry boundaries" (watersheds) and "wet boundaries" (the rivers), nor with the attempts to combine them. In large part, the general theory seems to have been supported by the concepts of natural philosophy which had a marked influence on nearly all the geographers at the end of this period. Indeed the expression of this philosophical concept in the geography of the time was taken up by political leaders--at least when it suited their particular purposes. On this basis Dalton justified the expansion of revolutionary France to the Rhine [1, 258 f.; cf. Spörer's long footnote on similar political arguments of a later period, 68, 363].

While the concept of natural divisions of the lands was effectively introduced as early as 1773, by Gatterer, the consideration of each of the divisions as in itself a natural unit does not appear clearly until 1805, in the work of Hommeyer [Bürger, 11, 7-12]. With Hommeyer, as with his predecessors, this unity represented perhaps little more than the unity of the landforms. The concept of a composite unity--the integrated total of all the phenomena of an area into an individual unit, distinct from those of neighboring areas was stated emphatically by Zeune, 1808-11, and by Butte in 1811. For Butte the individual lands and districts were "organisms" which like any organism included a physical side--inanimate nature, and a psychical side--animate nature including man. "The unit areas (Räume) assimilate their inhabitants" and "the inhabitants strive no less constantly to assimilate their areas" [1, 231].

Other writers of the period, however, did not accept these concepts without question. Rühle von Lilienstern immediately, 1811, opposed Zeune's theory with the impracticability of establishing definite "natural regions." A decade later both Wilhelmi and Selten urged that the boundaries of unit areas were not to be determined on the basis of any one kind of phenomena and recognized the difficulty of establishing definite limits where the boundaries of many different kinds were each gradual rather than sharp and, taken together, failed to coincide. "Nevertheless," said Wilhelmi, "the forms of nature are clear and distinctly separated as soon as we regard the particular constitution (of the total of all factors in any area) in its full form, rather than the boundary or transition zone" [from Bucher, 51, 89; also 1, 245].

The most thorough-going criticism of these concepts was made by A. L. Bucher in 1827. In a discussion fifteen years earlier, he had, with some hesitation, adhered to the concept of natural boundaries and apparently also--though this point is not clear--to that of regions as natural units. Rühle's critical discussion of Zeune, however, had made a strong impression upon him, his own efforts to establish natural divisions between parts of the continents had made him still more sceptical of the theory, and the discussions of Selten and particular Wilhelmi revealed, he thought, its fundamental fallacies--even though neither of them had arrived at that final conclusion.⁷ His concern over this and similar problems in geography led him to make a critical examination of nearly forty geography texts published in Germany in the first quarter of the nineteenth century. This remarkable study, whose value was at once perceived by Berghaus [55, editor's footnote, p. 516], merits a high rank among critical works in geographic methodology. Wisotzki rightly gives its discussion a prominent place at the conclusion of his survey of the pre-classical period. The arguments with which Bucher attacked the concepts of natural boundaries and natural regions leave little for the present-day writer to add.⁸ Indeed it would hardly seem too much to say that if Bucher's study had not been so generally overlooked by recent German as well as American geographers, much of the discussion of regions as natural concrete units in a later section of this paper would be superfluous.⁹

As the culminating expression of methodology of regional geography of the pre-classical period, Bucher's study represents in a sense a dead end. The long search for the true "natural divisions" he convicted as futile from the start, but he would not return to the use of political divisions. Rather he suggested that it was not necessary to establish either boundaries or divisions; regional studies were needed only for special purposes for which the areas concerned could be arbitrarily bounded in any convenient way [84-93]. The essential work in geography was concerned with the systematic studies of individual categories of phenomena, each however in relation to the earth. As Bucher's final study falls well within the "classical period" we will have occasion to consider these conclusions again in that connection.

Certain other features of the geography of the pre-classical period may be more briefly noted. The impression which the view of an area makes upon the spirit of the beholder was emphasized by Hommeyer, who considered the study of the aesthetic-geographic character of the Landschaft as a part of geography. Hommeyer, however, was one of the first to introduce confusion in geography regarding this term, as he specifically defined it as a portion of territory, intermediate in size between a Gegend and a Land [1, 211 f., 220]. More important in the development of the aesthetic description of nature, according to Humboldt, were Rousseau, Buffon, Bernardin de St. Pierre (in Paul et Virginie), Chateaubriand, Playfair, and Goethe, and, among geographers, particularly George Forster [43, 13; 52, 27; and 60, II, 65-75].

It was a natural consequence of the manner of development of the new science of geography that its proponents should see it closely related to history, but they insisted that geography should not be "the handmaid of history," but rather that the two were "sister subjects." The comparison of the two fields was most clearly stated by Kant, as we shall see in a later section. The extent to which geographical studies should include historical treatment was discussed by Rühle, in 1811, and, notably, by Bucher in 1827 [51, 237 ff.].

Though the new geography found its basis in physical rather than political features, it did not exclude the specific problems of political geography, as we understand the term. On the contrary, one of its claims was that it provided a firmer foundation for the study of that field than had a geography which constituted little more than political and historical geography. In particular, the drastic changes in the areas of states during this period constantly engaged the attention of geographers, and as we have noted, the concept of "natural boundaries," first developed as a framework for physical geography, was carried over--all too simply--both into political geography and political practice [l, 258-9]. One byproduct of this was the further confusion of the term "natural boundary" to include linguistic boundaries [240], a concept which recently has again been emphasized--not without inconsistency, to be sure--by one of the rulers of Europe.

Although the new science of geography sought its foundations in the more permanent physical features of the earth, it was characteristic of its time in considering these primarily in relation to man.Throughout the period most writers, consciously or unconsciously, followed Kant in regarding physical features in terms of their ultimate importance to man. Müller, in 1785, and Kayser, in 1810, specifically emphasized the study of the earth as "the dwelling place (Wohnsitz) of man." On the other hand, one can observe a swing away from this viewpoint in those writers like Hommeyer who emphasized the description of the earth forms--in terms of pure description, without explanation--as the sole content of "reine Geographie"; all other features might be considered in applied geography [Plewe, 8, 19].

One other concept of geography expressed at this time was that it was a study concerned with the "Where" of things. Stated by Lindner in 1806, it was vigorously opposed by Ritter in the same year.

Finally, it may be added, the students who were consciously endeavoring to construct a science of geography, repeatedly emphasized the importance of two cardinal principles of any scientific work. In contrast with the blind dependence on authority, characteristic of earlier periods, and the a priori construction of systems of supposed facts about the earth that characterized much of the work of the eighteenth century, many students toward the end of this period emphasized the importance of determining first the actual facts. Likewise they wrote repeatedly of the need to indicate specifically the sources of information [1, 255 ff.].

In large part, however, the new science of "reine Geographie" represented the expression of purposes and hopes rather than of accomplishment. In spite of the insistence on establishing facts, the geographers of this period, with few exceptions, continued to set up a priori systems of facts without putting them to the test, and many of these concepts we still inherit--or have constructed anew. In the century and a quarter since Butte spoke of areas as organisms no geographer, so far as I can find, has seriously attempted to establish the organic character of any single region, and yet, as we will see, current textbooks of the most modern type repeat this a priori assumption. In the century since Bucher's scientific integrity led him to announce his failure to establish a division of lands into natural regions, no one, so far as I can find, has seriously attempted to show the error of his methods or argument, or to produce what he found impossible; nevertheless the concept of the region as a definite unitary entity is prominent in the writings of a large number of geographers today.

If the geographers of the late eighteenth century developed the greater part of the theoretical concepts of the new science of geography, the transformation of their ideas, demands, and wishes into facts" was largely the work of Alexander von Humboldt and Carl Ritter.¹⁰ It is an extraordinary that modern geography in all lands should owe so much to two men living at the same time in the same country--for over thirty years in the same city.

This situation was not merely the result of a coincidence. Both of the "founders" of geography depended in large part on their predecessors whom, of course, they had in common. Furthermore, Ritter repeatedly asserted--and no student of his work has questioned the fact--that in many respects Humboldt, the elder by ten years, was his teacher. Peschel rather emphatically endorses Ritter's statement that, without the work of Humboldt, his own work could never have been produced [66, I, 324; 49, I, 54 f.]. Though we may find some exaggeration in the statement, its essential truth is testified by the innumerable places where Ritter echoes the concepts of Humboldt,¹¹ as well as by his hundreds of specific references to the latter's writings. That these are not merely the meticulous notations of a pedantic scholar is testified by Humboldt himself. In the introduction to his final work on Central Asia, he refers to the earlier publication of a series of itineraries and observes that "the terminology that I proposed for the mountain systems of Asia, and the geologic views on their mutual independence, have been discussed with great superiority by my old and illustrious friend, M. Carl Ritter, in his great work on Asia." [59, I, xlviii].

Although Humboldt made repeated use of Ritter's work in his own writings--the citations of Ritter in Asie Centrale must run well over a hundred--his general concepts of geography were evidently well formed before he met the younger student and before the appearance of Ritter's first geographic writings. Humboldt had been trained neither for geography nor for science, but rather for governmental service--through courses in technology, economics, history, philology, etc. His dominant interest however was in the study of nature--as a mine inspector he was concerned to study subterranean vegetation--and a close relationship with George Forster turned this interest in the direction of geography. In particular a trip which they made together in the Rhineland and to England, in 1790, stimulated Humboldt's ambition to undertake scientific travel and introduced him to Forster's methods of geographic study [47, I, 41; 60, II, 72; cf. 20, I, 95-108; 8, 23-4; see footnote 5]. It may well be that Humboldt's concept of geography had already been influenced, indirectly, by Kant [20, I, 46, 50 ff.].

During his short period in governmental service Humboldt published a number of scientific papers, including one or two in geography, but his great work, of course, came as a result of his travels in tropical America, in 1799 to 1804, as well as the later trip to Central Asia, in 1829. It would be an error, however, to suppose that his geographic work was limited to his own observations, or even to the lands which he had observed. Of over sixty years of active geographic study, his explorations hardly totaled five years. (To be sure much of Humboldt's time in the latter part of his life, when he was a very active privy councilor to the king of Prussia, was occupied with state problems.) Many of Humboldt's studies, even before the Kosmos, extended over the whole world and depended on the utilization of the work of countless other explorers and scientists. Indeed, during the greater part of his career Humboldt was as much a geographer of the study and library as was Ritter.

Although Humboldt's work had an important influence on Ritter throughout his career, Peschel was unquestionably exaggerating when he stated that their first personal contact, during several weeks in 1807, was the critical experience that definitely turned the younger student to geography. Ratzel has shown that neither this statement--frequently repeated by later writers after Peschel--nor the more general inference that Ritter came into geography from history, is supported by the facts of his life [26, 423, 389 ff.]. Probably no student of his time--and few since--have been so specifically trained for geography as was Ritter.¹² From very early childhood until he went to the university, he lived in a school, in Schnepfenthal, in which education followed the precepts of Rousseau and Pestalozzi: knowledge of the world was acquired by direct observation of nature on walks and longer trips. Throughout this period the youth was under the particular personal care of J. C. F. GutsMuths, a teacher whose major interests were in nature study and in geography, and who himself made some contribution to the development of geography [Ratzel 26, 383, 388 ff.; Plewe 8, 39]. At this time, geography was Carl Ritter's favorite study, and in the following university years the natural sciences commanded his chief attention together with history and theology. These studies did not end when he left the university: for nearly twenty years as tutor in a private family in Frankfurt, he had both the need and the opportunity to continue the study of natural sciences, geography, and history, and to make numerous field trips in the Rhineland, Switzerland, etc., making careful, detailed observations. Likewise he was able to come into close relations with many eminent scholars of the day, of whom the most significant for him were the anatomist, Sömmering, the geologist, Ebel, and the leader of educational reform, Pestalozzi.

In sum, we may conclude that Ritter's preparation for research covered an extremely wide field, in which on the one hand natural sciences, and particularly the observational methods of nature study, predominated, but in which, on the other hand, his interest was increasingly in human problems, i.e., history. Geography, as he read it in the writings of Herder, a follower of Kant, maintained the connection between his earlier field of studies and his ultimate interests [cf. Schmidt, 7, 78 f., and 41-4]. Although his meeting with Humboldt, in 1807, undoubtedly stimulated his interest in geography, for the very reason that Humboldt had demonstrated so clearly the importance of earth conditions to man, for at least ten years thereafter Ritter was still uncertain whether to continue his geographic work or to shift to history.

His first publications--overlooking a series of geographic studies for elementary students--dealt with the geography of Europe. These appeared between 1804 and 1807 and made a considerable impression in Germany. One of these, which was also enthusiastically reviewed in France, was a brief study in the systematic geography of Europe, based on six maps, with explanatory text, of individual categories of phenomena. Appearing some years before Humboldt published similar maps, it constituted, according to Plewe, "the first physical atlas" [8, 30; cf. Kramer, 24, I, 255]. Though all these geographical studies appeared before the celebrated meeting with Humboldt, one may presume that Ritter had already been influenced by the latter's writings, since he wrote at that time that he "had devoured all his published works with a kind of ravenous appetite" [24, I, 167].

During the next few years Ritter prepared a more complete systematic geography of the world, a "Handbuch der physischen Geographie," which he intended to be the completion of his geographic work and a bridge to studies in history, to which he wished then to devote the rest of his life. For some reason however--possibly because of the criticism offered by the geologist, von Buch, who read the manuscript at Ritter's request--he decided not to publish it. Parts of it however were utilized by various colleagues to whom he freely offered it, so that the work appeared in a number of texts of the time, in some cases with the original authorship indicated, in others not [24, I, 205-7, 258-68].

During the same period Ritter's studies on Asia led him into the historical field, in which he prepared a lengthy study in ancient history (published later, 1820)--according to Ratzel "the only work of Ritter's of exclusively historical character" [26, 410]. Apparently, however, he contemplated shifting entirely to history. To fulfil his geographical work he outlined a project which should cover the world by regions--originally in a four volume Erdkunde. This would then supply the sound foundation for historical studies. Working at the University of Göttingen--where at the same time he took further courses in history, but also in mineralogy--he completed the first volume, on Africa, according to plan, but soon found that the enormous amount of material that he collected on Asia could not be handled in one volume, so that he issued the second volume as but a part of the study of Asia. The appearance of these two volumes, in 1817 and 1818, made a tremendous impression in the academic world; almost at once, says Marthe, Ritter was recognized "without question as the reformer of geography, as the master who first made that field into a science" [25, 6; see also Kramer, 24, I, 377 f., 404 f.; Ratzel, 26, 689 f.; Schmidt, 7, 86].

Ritter's position as the master of academic geography was therefore definitely established before he had held a single academic post. After one year as professor of history and geography at the gymnasium in Frankfurt, the efforts of various officials in Berlin--including Wilhelm v. Humboldt, the brother of Alexander [24, I, 436, 447]--were successful in arranging for a double position for Ritter, in the military college and at the university. In 1820, Ritter went to Berlin as the first university professor of geography. At just what time he recognized that his plan of "finishing up" on his geographic work in order to go on to history was unattainable is not clear; in any case he continued to work in geography in Berlin during the remaining 39 years of his life, although it may be said that in the latter volumes of the Erdkunde, his interest in history became more and more pronounced.

Following Humboldt's final shift of residence from Paris to Berlin, in 1827, the professional relations between the two increased steadily. According to Kramer, who lived for some years in Ritter's house, few days passed, at least in the latter years, in which--if they were both in Berlin--communications of greater or less importance did not pass between them [24, II, 85].¹³

This is not the place to attempt an appraisal of Ritter's contribution to geography, nor should such an attempt be necessary. No geographer in history had his work discussed so repeatedly, and often critically, as Ritter. The studies referred to in the following paragraphs are no mere obeisances to a great "traditional" figure. Except for the quotations from Humboldt, they are taken from critical examinations of Ritter's work by authors who did not hesitate to point out its weaknesses and the extent to which it failed to live up to his stated purposes. Nevertheless those students who have thoroughly examined his work appear to agree unanimously with Humboldt in recognizing its "masterly" character. The more important critical studies of Ritter's work are those of Peschel, Marthe, Ratzel, Hözel, and Wisotzki [66, I, 336 ff. ; 25; 26, 405-28; 27; 1, 267-323].

In view of the real differences between Ritter and Humboldt, both in personal character and, as a result, in their work, but particularly because of the supposed opposition between their points of view in geography, it is appropriate to note Humboldt's opinion more fully. The earliest expression that I have found is in the lecture course that Humboldt gave at the University of Berlin, 1827-28, Humboldt did not limit himself to the laudatory phrases which the occasion called for when he described Ritter's Erdkunde--at that time only the first two volumes had appeared--as "the most inspired work of this kind (comparative geography), that our century has yielded; it is the first work in which is presented the influence which the surface-view has had on the peoples and their fates" [52, 14 f.]. In a personal letter to Ritter a few years later, 1832, he expressed his great admiration of the work on Asia, adding that it gave him great pleasure to sing its praises to the king and all at the court, telling them that "such an important work had not appeared in thirty years" [24, II, 120]. Humboldt used his influence at the court to have Ritter relieved of some of his heavy duties, presumably at the military college, in order that he might complete the work on the Erdkunde [24, II, 32 f.; 20, II, 127]. That Humboldt did not alter his opinion is indicated by numerous statements published in later writings.In the introduction to his Central Asia he finds Guigniaut's verdict correct: "one of the greatest and most magnificent monuments raised to science in our day" [59, I, xlviii f.]. In the first volume of the Kosmos (1845), he speaks of "Carl Ritter's great and inspired work" which "has shown that the comparative geography attains thoroughness only when the whole mass of facts that have been gathered front various zones, is comprehended in one view, is placed at the disposal of the integrating (combinierenden) intelligence" [60, I, 18]. Later in the same volume he refers to Ritter's work as the fulfillment of a part of the plan of Varenius, which--in the earlier period of scientific development--Varenius himself had not been able to accomplish: "It was reserved for our time, to see the comparative geography cultivated in masterly fashion, in its widest compass, indeed in its reflex on human history, on the relations of the form of the earth to the direction of the characteristics of peoples and the progress of civilization" [60, I, 60, with footnote reference to Ritter].

The new direction which Ritter was endeavoring to bring into geography is explicitly stated in many writings from 1804 on.His first principle, frequently repeated, was that geography must be an empirical science, rather than one deduced either from rational principles--from philosophy--or from a priori theories of "general geography." "The fundamental rule should assure truth to the whole work," he wrote of his Erdkunde, "is to proceed from observation to observation, not from opinion or hypothesis to observation" [49, I, 23]. (In contrast to deductive reasoning from a priori theories as described on page 48.)

The first major plank in Ritter's reform of geography therefore was to call a halt on all attempts that led from theoretical considerations to systems of earth forms. He was among the first to show that the theory of a continuous network of mountains was opposed by the records of observed facts and that, likewise, there was no general correspondence, as often assumed, between the crest-lines of the mountains and the divides of drainage basins. Although the teleological view of the universe that Ritter had received through Herder from Kant provided the background of purpose of his study, he was not, like Kant, a philosopher who required geography in order to establish theories of world knowledge [cf. Adickes, 13, 75 f., 190 f.; Gerland, 12]. Convinced that there were laws governing the relation of human and non-human phenomena on the earth, Ritter was in no great hurry to establish them, but rather felt that, if he could bring together all the facts and relationships observed in areas, these would make it possible to state such laws. Attempts to develop these laws first--his own, perhaps, as well as others--had proven unsuccessful: "we must ask the earth itself for its laws" [49, I, 4].

Ritter was therefore the first great opponent of what may properly be called "arm-chair geography." On the other hand, as a scholar and teacher, rather than explorer or field worker, as Hettner observes [2, 310], Ritter represented the geographer in library and study: the observations used in his writings were not obtained directly, but rather from the works of others who did observe at first hand.Not that he did not travel and observe, as is often erroneously supposed; on the contrary he traveled much throughout his life, making detailed observations in many parts of Europe [24, I, 271-331, II, 64-83]. But since he unfortunately became so deeply involved in his studies of Asia as never to commence the volumes on Europe, the detailed observations which he made on his own trips had only the value of enabling him to "understand what others had observed."

In basing his work on observations of other students, Ritter was not satisfied with the limited amount of material which had sufficed for most of his predecessors. In order to achieve the highest degree of accuracy, his plan was to bring together on each point in question the greatest number and variety of trustworthy witnesses, of all times and from all peoples, so that their testimony might be seen together, whether in agreement or for comparison [49, I, 23]. The material was by no means limited to the works of travelers, whether scientific or otherwise, but was drawn also from the works of a large number of specialized scientists, as one may observe by glancing through the introduction or by noting the footnotes in any of the volumes.

It seems more than probable that Humboldt had Ritter's methods of work, as well as his own, in mind, when he discussed the use of sources in his lecture course given during the winter of 1827-28, at the University of Berlin. "What an individual observer can see, is naturally small in comparison with that which has been observed during so many centuries. If it is therefore very important to gather the observations, it is likewise necessary to be engaged in some part of the sciences of nature, for only so can one learn to understand that which others have observed. For those who can spend the great amount of time and energy necessary, the most important method is the study of all travel descriptions, of all the individual treatises, for only out of the special cases can the general be recognized..." [52, 26].

One cannot glance through the pages of Ritter's presentation of Asia without being impressed by the great number and variety of the sources upon which he depended: the works of contemporary scientists like Buch and Humboldt--whose works are no less honored by use than by praise--the reports of earlier travelers, whether non-scientific travelers or those of whom Humboldt complained that, while "well prepared in the isolated branches of science," they rarely had "a sufficiently varied knowledge" to study phenomena of different categories in interrelation [47, I, 4]; and in addition a great host of works of all kinds from the pens of medieval and classical writers. In large degree, these were the same kinds of sources, in many cases the same works, used by Humboldt in his writings. Although the very quantity involved made it impossible for Ritter to complete his original purpose of covering the whole world--and indeed deflected him from his particular aim of correlating the human and physical facts established--he at any rate could claim to have assembled for comparison all the evidence available to establish the facts. The first of many to testify to his success in this regard was Humboldt, with particular reference to the area with which he himself was personally familiar, Central Asia. Shortly after Ritter sent Humboldt the first volume on Asia, in its second edition of 1832, Humboldt wrote to him that he could find no words in which "to express the real admiration with which I am filled by your gigantic work on Asia. For the past two years I have been most earnestly engaged with Inner-Asia, with the use of all sources, but how much has become clear to me only in the three days in which, without break, I have been reading in this work. You know everything that has been observed for centuries, with your particular sagacity you arrange everything together, from much-used material you gain ever new and grand views, and you reproduce the whole in the most desirable clarity" [24, II, 120].

It was not, however, Ritter's purpose merely to assemble a mountainous accumulation of information. The enormous multiplicity of observations must be organized according to the chorological (räumliche, or spatial) principle, which he considered both in a horizontal, and--with reference to elevation--in a vertical sense [49, I, 24]. This did not mean, however, simply an accumulation of facts concerning each area; rather he wished to show the "coherent relation" (Zusammenhang) in terms of cause and effect, of the different features [24, I, 250 f.] and the "formation of the multiplicity of features into the essential character of an area" [8, 32]. By proceeding "from the simple to the compound," he wished to establish the totality of interrelated features as the distinctive character of each area [49, I, 24]. We are but paraphrasing any number of his statements if we say that he was concerned with reducing to order observations made in nature [49, II, xv ff.].

In these respects Ritter's purposes were not so sharply in contrast with those of his predecessors as he believed, but the great difference was in the extent to which he put these into practice--even though this was by no means complete.In certain other respects his position was in clearer opposition. Though he followed the movement of "reine Geographie" in using "natural lands" rather than political states, he objected to the overly-simple, and in fact impossible, procedure of dividing only by mountain ridges, or the equally simple division by drainage systems. According to his theoretical statements his divisions were made in terms of relief (in particular he divided the upper, middle and lower portions of stream basins), but in practice he also recognized other factors, as in his separation of the Sahara and in its subdivisions in terms of climate, vegetation, etc. [49, I, 959 ff.]. In this, he was consistent with his fundamental principle, that we should not evolve a theory of natural divisions of the lands, but must always look to nature for its principles [49, I, 4].

He reminded his contemporaries, in their intense argument over the relative values of "wet" and "dry" boundaries, that the mere drawing of boundaries was but a means toward the real purpose of geography, the understanding of the content of areas. This view he emphasized by a phrase which, though not clear in itself, was perhaps sufficiently clear in the context in which he used it as to justify its frequent repetition in later decades, so long as the context was not overlooked.Geography, he suggested, is the study of "der irdisch erfüllten Räume der Erdoberfläche"; i.e., the areas of the earth are not to be studied in themselves, as mere divisions of the earth's surface, but neither are the objects that are found on the earth surface to be studied in themselves--in geography--but rather the areas of the earth surface are to be studied in terms of the particular character resulting from the phenomena, interrelated to each other and to the earth, which fill the areas [50, 152].¹⁴

In all of these respects there was no essential difference between Ritter's view of geography and that of Humboldt. Since Ritter had expressed many of these views as early as 1804 [24, I, 250], one may question Peschel's statement that it was in their first personal contact in 1807, that Ritter "was brought to see clearly the great function of geography--to show the unison of human social phenomena with the complex of natural forces in the locality" [66, I, 324, 341; see also Döring, 22, 160-3]. Ritter may however have acquired his ideas from Humboldt's writings, with which he was already familiar, and the lengthy conversations with Humboldt over several weeks may well have emphasized and clarified his concepts [24, I, 165-7]. Furthermore Peschel is simply echoing many of Ritter's statements when he says that the latter's work in relating historical phenomena to physical phenomena would have been impossible had not physical geography been advanced so notably by Humboldt: "a Carl Ritter could not have preceded an Alexander v. Humboldt, but could only follow him" [66, I, 324].¹⁵

On the other hand, the great difference in temperament and in general outlook on life gave a very different color to the work of the two founders of modern geography, and in the works of their followers this difference became fundamental. It is natural enough therefore that later students should tend to emphasize the differences between the two men and even to find an opposition between them of which neither of them were conscious--familiar though they were with each other's work. Interesting and dramatic as such a contrast may be, historical objectivity does not permit us to exaggerate differences that may be superficial into a fundamental contrast. (See Supplementary Note 4)

(Subsequent to writing this conclusion, I find that Ratzel had made much the same statement, emphasizing the many respects in which Humboldt's geographical work, even in its philosophy, was similar to that of Ritter [26, A. d. Biog., 690 f.].)

In fairness to all varieties of critics of Ritter it must be said that in his voluminous and variegated writings one may find apparent justification for almost any possible judgment. If at times he wrote as a scientist, at other times he wrote as a philosopher--indeed it did not occur either to Ritter or to Humboldt that science and philosophy could be completely separated--and at other times Ritter wrote as a theologian. Furthermore, Ritter was unfortunately far from careful in his selection of terms, "he did nothing to confine the fluid and ambiguous quality of words," "strict logic was not his forte" [Plewe, 8, 28]. Any number of writers have debated what Ritter intended in calling his Erdkund, "allgemeine vergleichende" (general comparative) and the net conclusion appears to be that we are not to look for any specific meaning in these terms, but merely to understand that he had certain general purposes in mind that he wished to contrast with those of his predecessors [see Hettner, 2, 310; and Plewe, 8, 28 ff.].

Ritter is therefore a writer who cannot be judged, either in his work or in his methodology, on the basis of a few striking quotations pulled out of their context and out of their relation to the ideas of his contemporaries. A view of Ritter constructed in this manner has the effectiveness, but only the degree of truth, of a caricature. We need not however make an exhaustive study of Ritter's work, but may base our appraisal on the many thorough studies of that kind made by previous students, in each case, however, referring their findings to the original work. The latter must include his actual geographic work as well as his methodological studies, since the methodology is expressed no less in the actual work itself.

In particular, one must avoid the error of attempting to determine the character of the geographic work of either Ritter or Humboldt--or of any other writer--from their own statements of ultimate purposes. The fact that Columbus set out across the Atlantic for the purpose of finding a shorter route to the Indies, and apparently died still believing that he had found it, does not cause us to judge his importance in history in terms of the establishment of connections with the Indies--to that he contributed little more than a permanent confusion in geographic and ethnographic terminology.

Writers of a later period, living in a different philosophical atmosphere, attempted to discredit the importance of Ritter's work by attacking his philosophical concepts, in particular his pietistic, teleological view of the universe. Commonly this is done under the naïve assumption that the writer has a purely "scientific" concept, in contrast with a philosophical concept of the universe. We cannot here discuss this elementary philosophical problem, but merely point out that every scientist has his philosophy of science, and that that philosophy is neither science itself nor is it the product of science. That Ritter recognized this distinction may be seen particularly clearly in his discussion of Fröbel's criticism of his work. To Fröbel's objection to his teleological view--which was not an argument but simply a different philosophical assumption of science asserted without foundation--Ritter replied properly in philosophical terms [55, 517 f.].¹⁶Though we may reject his philosophy we call respect his statement of it. On the other hand, in reply to the attack on his scientific methods and concepts, he limits himself to scientific terms.

The proper question to ask in criticizing the scientific works of any writer--as Plewe among others has noted--is not: what particular philosophy of science is exposed by these writings, but rather: to what extent does the writer's particular philosophy affect his work. One may have a teleological philosophy--as did Leibnitz, for example--and nevertheless still produce genuinely impressive research [8, 70 f.]. Though many of Ritter's followers, like Guyot [64], depended on the teleological factor to explain geographic details, competent critics agree that Ritter seldom was guilty of this procedure, which in fact he vigorously condemned [Marthe, 25, 21 f.; Wisotzki, 1, 297-304].

In general, we may say that the teleology in Ritter's geography was an attempt to interpret philosophically that which science could not explain. Many of the statements at which the scientific critics of the late nineteenth century pointed in holy horror would stand comparison with current utterances of a number of eminent physical scientists who find themselves in a similar position. In Ritter's case, there were three fundamental facts of geography for which science had no explanation--and presumably still has none--namely, the uniqueness of tile earth in the universe, so far as we know it; the earth as the home of that unique creature, man; and finally--the fundamental explanation of a host of geographic facts--the differentiation in character among the major land units of the world.

In our day, to be sure, it is more common for scientists to express their philosophical views of problems presented by their scientific work in separate publication; the intellectual world of a century ago did not call for this physical separation, nor can we regard it in itself as a major difference. "Even though a spirit turned toward the eternal breathes through Ritter's presentation," concludes a modern student, Schmidt, "even though his highest enthusiasm is fired by his religious Weltanschauung, nevertheless in his research, he in no way proceeds from preconceived opinions; his scientific procedure was directed throughout on temperate, purely factual comprehension of the facts and their relations...Ritter strove in the knowledge of the earth for a comprehension of the divine world plan in no other way than the natural scientists pursue the thought of evolution" [7, 85].

Although most critics would add some degree of qualification to Schmidt's judgment, it seems fair to conclude that the disturbance caused in the modern scientist's mind by Ritter's teleological expressions is far out of proportion to their importance in his work. An illustration is offered in the case of his treatment of the continents, specifically in two lectures given before the Berlin Academy of Sciences, one in 1826, the other in 1850, republished later together [50, 103-28, 206-46]. They constitute some 65 pages in which the author describes the size, form, construction, and climatic conditions of the different continents, seeks to construct "a law of the arrangement" of the parts of each continent to the whole [220], and particularly endeavors to show how these conditions have determined the development of the peoples of the different continents at different periods of history. One must look with a very careful and critical eye to find any statement that would disturb the scientific determinists of the late nineteenth century. Near the beginning of the first lecture, to be sure, Ritter expressed his belief that the earth had been planned as "the temporary nursery of the human species"[104] and a sentence at the end of the second lecture suggests that a part of this plan was the particular form and position of each of the continents, which had led--as his detailed discussion had demonstrated--to the particular function that each had played in the course of world history [243].¹⁷Those who can overlook such expressions of belief will, like Humboldt, test the essays in terms of their scientific content.

Humboldt's comment in this connection also throws light on his interest in human geography. Noting that he himself had long insisted on the great influence that the form of a continent exerts on its climate and vegetation, he discusses Strabo's effort to show the relation of the articulate and peninsular form of Europe to the development of its civilization, and adds: "In our time M. Ritter has developed with great wisdom the analogies, in physical and political characteristics, that the three peninsulas of Asia offer to the three peninsulas of Europe. The two groups offer centers of culture of very distinct physiognomy" [59, I, 67 f., footnote].

Nevertheless there were certain respects in which Ritter's philosophy did affect his work, and even more that of his followers. His purpose led always from the individual facts toward a "Whole" of all phenomena; rather than to trace the inner relations of phenomena to a "last cause," his purpose was to find the coherence of forces in a Whole, and thus ultimately to indicate the purpose of the Whole [Wisotzki, 1, 304]. This purpose must surely be found most particularly in the highest of earth creatures, the only one who could conceive of an organization of the Whole, namely man (it was logically false, he replied to Fröbel, to suggest that it could just as well be found in cows [55, 518]). His consideration of the earth therefore logically centered on man.

This view, which may be found in his first geographical publication, in 1804, is expressed particularly in the statement of purpose with which he introduced his Erdkunde: "to present the generally most important geographic--physical conditions of the earth surface in their [or possibly "its"] natural coherent interrelation (Naturzusammenhang), and that (the earthsurface) in terms of its most essential characters and main outlines, especially as the fatherland of the peoples in its most manifold influence on humanity developing in body and mind" [49, I, v].¹⁸

While this concept of geography unquestionably places a major emphasis on the earth as the home of man, it does not restrict it to that particular view. "Independent of man, the earth is also without him, and before him, the scene of the natural phenomena; the law of its formation cannot proceed from man. In a science of the earth, the earth itself must be asked for its laws" [49, I, 4]. The first volume of the Erdkunde gives detailed consideration to areas in Africa of apparently no concern to man [cf. Plewe, 8, 70 f.]. In the later volumes, however, Hettner finds that Ritter concentrated his attention on man to such a degree as to forget his purpose of establishing the total association of nature and man [161, 87; cf. Marthe, 25, 18; Ratzel, 26]. Nevertheless one of his major accomplishments, according to Marthe, was to establish the physical foundations of geography in detailed study of what we would term regional landforms [25, 24 ff.].

Further, Ritter's particular philosophical view--but also perhaps his failure to submit concepts to rigid logical examination--permitted him to further the traditional concept of regions as "natural divisions" of the earthsurface. That Ritter's natural philosophy has continued to affect, in this respect, the views of geographers who would deny that philosophy has been repeatedly suggested by Hettner [in 1908, 300, 7-13; repeated, 161, 299-306].

Finally one may conclude that in his teaching, in contrast to his writing, Ritter's teleological views were more strongly impressed upon his students. There is some evidence for that in the lectures published shortly after his death [61], and far more in the performance of his students.

Unquestionably the modern scientist feels more at home in the writings of Humboldt than in those of Ritter. Peschel, though deducing from certain of Humboldt's statements that he could not have been a materialist, honors his avoidance of the confusion "not uncommon with the materialists as well as with their opponents, of presenting scientific investigations in some form of religious light" [66, I, 305]. On one point, to be sure, Peschel finds that "the noble heart of Humboldt appears to have a bit corrupted his critical sense," namely in regard to race questions. Humboldt vigorously maintained the theory of the "unity of the human race," a concept that "came to prevail first through Christianity" and could be ignored on the slave-market only because of "the degeneration of Christianity by great wealth." The opposite "hypothesis of racial gradations among men" he characterized as "not only unkind (lieblos) but also false"[52, 183-4; 60, I, 385 f. and footnote]. Although Humboldt's thesis would find more support among anthropologists today than that on which Fröbel as well as Peschel based their criticism, the criticism itself was essentially justified: his views on this (question were strongly influenced by non-scientific considerations. In his youth, Humboldt had first found intellectual stimulus in what seemed to Goethe, Forster, and many others as the "barren environment of Berlin," almost exclusively in a small circle of Jewish intellectuals. From them he had learned of Lessing and Kant [according to both Löwenberg and Dove, 20, I, 40-49; II, 292 f.]. He himself described the impressions made upon him by the degrading scenes of the slave-market, which he first witnessed in tropical America [47, I, Bk. II, Chap. 4]. Both in political circles and in his scientific writings he repeatedly and vigorously expressed his views on the race question. In his master work, the Kosmos, he chose to "close the general presentation of the natural phenomena of the universe" with the following quotation from his brother, Wilhelm v. Humboldt: "We wish to note one idea which is visible in ever increasing validity through the whole of history..., the idea of humanity,...to treat the whole of humanity, without consideration of religion, nationality, and color, as One great closely related race, as one Whole existing for the attainment of one purpose, the free development of inner powers"[60, I, 385 f.].¹⁹

In common with most other students, Peschel fails to note the extent to which Humboldt's philosophical views find repeated expression in his scientific writings, nor does he note a certain similarity between those views and the concepts which Ritter expressed, commonly in a more religious tone. Geography was not, for Humboldt, a field studied as an end in itself, but rather as the means of comprehending "the harmonious unity of the cosmos" as a "living whole," "a unity in multiplicity" [60, I, 4 ff.]. "Insight into the cosmic organism (Weltorganismus) creates a spiritual enjoyment and an inner freedom which even under fate's hard blows can not be destroyed by any external power [44, 32; cf. 60, II, 89 ff.].

For Humboldt, this unity of nature was no teleological, anthropocentric unity-he wrote of the comprehension of "the inner, secret play of natural forces" [44, 32], where Ritter spoke of discovering the "divine secrets" [50, 228]-but was a balanced unity of the whole of nature, of which man was a part [note the many quotations in Döring, 22, 18, 37 ff., 59]. "In the forests of the Amazon as on the ridges of the High Andes, I perceived how, animated by one breath from pole to pole, only One Life is poured out in stones, plants, and animals, and in the swelling breast of man" [from a letter quoted by Rehder, 23, 136]. An important factor in the evolution of Humboldt's thoughts was his close personal relationship with Goethe [20, I, 187-201]. With him, and with the entire generation of the "romanticists" (in the widest sense), Humboldt shared the idea of an organic coherence of all phenomena; this was a common characteristic of both of the founders of modern geography [Plewe, 8, 49-51].

In other words both Humboldt and Ritter, in their philosophical point of view, were products of their time-in particular both were influenced by the thought of Kant and of Rousseau. To Ritter's religious nature the concept of a universe of order and law in which all phenomena of nature and man were interrelated required the assumption of some divine purpose and plan which the scientist should attempt to establish. Granted this assumption it appeared obvious that the ultimate purpose involved could not be simply the production of an intricate universal mechanism; the purpose could only be found in the life of the highest of the creatures of the universe, the only thing in the universe that was capable of recognizing either the order of the universe or a purpose behind it-namely, man. In Humboldt, on the other hand, the same philosophical concept found a responsive chord in his feeling for the aesthetic.rather than the religious. In the writings of Rousseau and St. Pierre, it was the descriptions of the "harmonies of nature" that appealed to him; through Goethe he received the concept of the "Landschaft" as conceived by the "romantic" movement [cf. Rehder, 23, 134-38].²⁰

Interesting as this comparison may be, between the philosophical viewpoints of the two founders of modern geography, it does not directly concern us here. The essential point is that both of them differed from their philosophical and literary predecessors in that both strove to demonstrate their philosophical concepts not by deductive logic, as in the case of Kant, nor by sentimental descriptions of subjective impressions of nature, as in the case of St. Pierre, but by objective descriptions of observations of nature. If there are exceptions in Ritter's work that we must overlook, the same is true, if in lesser extent, in the work of Humboldt [cf. Dove, 21]. Through out the writings of both students, the modern scientist finds repeated expressions of philosophical concepts that are now commonly excluded from such works But modern science, which has had bitter fights with religious philosophy, has never come into serious conflict with the aesthetic view of nature; perhaps for this reason Humboldt's expressions of his Weltan schauung are less disturbing to scientific readers today than are those of Ritter.

In general, neither Ritter nor Humboldt saw any conflict between science and philosophy. Ritter found that both geography and history "are directed toward the integration (die Combination und das Mass) of ideas and are therefore forced to philosophize" [50, 152]. Likewise Humboldt, a student of the works of Kant and Fichte, felt a need for "something better and higher, to which everything could be referred" and therefore greeted with enthusiasm the natural philosophy which Schelling had founded. In a letter to Schelling, in 1805, and in publication two years later, he expressed his conviction that "a true natural philosophy could not harm empirical research"; on the contrary "such a philosophy leads findings back to principles and is likewise the foundation for new findings" [20, I, 228 f.; 41, v]. Some years later, to be sure, when many students gave "pure thought" the preference over empirical studies and seemed to think that science could be produced by thought without research, Humboldt scathingly attacked this "mad" form of natural philosophy [20, I, 230].

That Humboldt did not associate Ritter's work with the "bal en masque" of a philosophy that despised observations and experiments and taught "a physics that one can pursue without wetting one's hands" is indicated by a discussion immediately following his praise of Ritter's work in one of his lectures at the University of Berlin: "The description of the universe provides materials for a proper natural philosophy, the foundations for which are sought in many different ways. I can not find fault with these efforts, although I would be inclined myself to go to work more empirically. In this natural philosophy we need only fear and avoid with difficulty: false facts, Empiricists and philosophers should not mutually despise each other, "for only bound together can we attain the highest goal" [52, 15].

For both Humboldt and Ritter, the concept of unity of nature presumed a causal interrelation of all the individual features in nature. The phenomena of nature were studied in order to establish this coherence and unity. For both it was axiomatic that the unity of "nature" included organic as well as inorganic, human as well as non--human, immaterial as well as material. The exclusion of any part would be not only arbitrary but would destroy the coherence and unity of the whole. In his first publication, in 1804, Ritter vigorously opposed the trend in "pure geography" that tended to over--emphasize, he thought, the "natiirliche Landschaft," insisting that geography must describe and explain all the present conditions of an area [24, I, 250 ff.; Piewe, 8, 32 f.]. Ultimately, to be sure, his predominant interest in man tended to emphasize the human element to such an extent that he actually failed to accomplish his fundamental purpose [Marthe, 25, 18; Ratzel, 26].

Humboldt's view of science was limited only by the universe outside his own mind, "Natur." Goethe, as well as Ritter, called him "an academy in himself" [20, I, 198; 24, I, 154]. Even if we eliminate those parts of his work that he did not regard as geography (Erdbeschreibung), so extensive and rich was his view of the field that almost every movement for some particular form of geography--from the geophysics of Gerland to the aesthetic geography of Banse can find its antecedents in Humboldt. None, however, may look to him as the precedent for excluding other parts of the field. "In the great enchainment of causes and effects, no material and no activity may be studied in isolation" [44, 39]. "The highest goal of all observation of nature is the knowledge of our own nature: and therefore we conclude our description with a consideration of the races of men" [52, 13, 182-90]. Likewise the outline of systematic geography in the Kosmos--called "physical geography" according to the usage of the time--culminates in the section on man [60, I, 376-86]. The consideration of man includes spiritual aspects as well as material, "as though the spiritual were not also contained within the natural whole (Naturganzen)" [60 , I, 69]. Moral and aesthetic problems therefore formed a part of his considerations [see 44, 24; and 47, I, 348-52]. (See Supplementary Note 5)

It is true that in many of Humboldt's regional descriptions one will find little concerning man and his works, but he himself explained that his travels in equatorial America had taken him chiefly to areas where "man and his products disappear, so to speak, in the midst of a wild and gigantic nature" [47,I, 32]. His purpose, in presenting "the general results that interest allenlightened men" is to describe not only the climate and "the appearance of the countryside (paysage), varied according to the nature of the soil and its vegetable cover," but also the "influence of climate on organic life," "the direction of the mountains and rivers that separate the races of men as well as the plant societies" and "the modifications found in the condition of peoples placed in different latitudes and in circumstances more or less favorable to the development of their faculties. I do not fear that I have increased too much the number of objects so worthy of attention: for one of the fine characteristics that distinguish present civilization from that of past times is that it has enlarged the mass of our conceptions so that we realize more clearly the relationships between the physical and the intellectual world; hence there has developed a wider interest in objects that previously were of concern only to a small number of savants because they had been considered isolated from each other according to an overly narrow point of view." [47, I, 14.]

In his descriptions of populated areas, Humboldt consistently strove to fulfill this purpose. Man, his culture and his works, are included as integral parts of his description and interpretation of nature. (In his studies of Central Asia, he was limited by specific restrictions imposed by the Russian government.)

Both Humboldt and Ritter carried the concept of unity of all nature into the consideration of individual areas, at least in what I have termed the"vertical" sense--i.e., that all the features of an area in their interconnections form a naturally unified complex, whether or not that is to be considered as a unit Whole or merely a part of the one natural Whole of the world. (See Supplementary Note 6) For Humboldt, every part of the world is a reflection of the unity of the Whole [60, II, 89]. Ritter's statements on this point are more emphatic; both however, were, as we have seen, following paths already suggested by their predecessors.

How was this unity to be studied? It hardly seems necessary to say that neither of these "founders" of a science supposed that the study of an area was to be confined to a consideration of it as a whole, or that it should begin with such a consideration of the whole. While both assumed from the start that all the objects and forces of nature (nature in the sense of the world outside our minds) were interconnected to form a Whole, both recognized, in theory as well as in practice, that this could be established scientifically only by investigating the individual, single features in their relation to each other, and building these up in their actual relationships to form the whole. The function of tile scientific geographer was to perceive these features, not separately, but in their interrelations so that he could thereby reproduce intellectually the unified whole that was nature [60, I, 4 ff., 51 ff.; cf. Plewe, 8, 39].

That the previous statements correctly represent Ritter's purpose and method is supported not merely by the views of many other students more familiar with his work, but appears clear from even a brief consideration of either his methodological discussions or of his work. To be sure, one may find sentences in his writings which appear to say the opposite. Thus a century ago the young student, Fröbel, extracted one or two of Ritter's sentences from their context so that it appeared as though Ritter attempted to use synthesis without preliminary analysis [54, 502, 504]--the "holistic" principle, as Leighly puts it, that "forbade the investigator to dissect a whole into its parts for fear of destroying its coherence and so its essential character" [222, 257]. That Ritter was not so naïve as not to recognize that synthesis presumed analysis can be read from the paragraphs that follow the quoted sentence on which Fröbel has based his argument [49, I, 3]; in deed every section of his work demonstrates the principle which he had definitely stated in his introduction, "that the procedure must lead from the simple to the compound, from the individual parts to the unit" [49, I, 24]. It was consistent with Ritter's personal character that he should have some hesitation in replying to the charge that his method "which builds up and combines" was one "in which analysis is unconsciously presumed" [Fröbel, 54, 504]. Many thought that he might have asked Fröbel to read again the pages referred to, but he contented himself with the suggestion that it might have occurred to his critic that it would not be news to the author of a synthetic system that synthesis must be developed from analysis [55, 515]. Later critics, indeed, questioned whether Ritter had actually succeeded in getting beyond analysis to synthesis.

The "horizontal" concept of unity of particular areas, as individual wholes, was also furthered by Ritter, though apparently with some qualifications. He did, to be sure, stress the individuality of the continents, which at times he referred to as "organs," but in his reply to Fröbel he insists that the terms are not to be taken literally, but in the form of analogies (just as, one may note, Fröbel had written of the systems of like phenomena on the earth as "organs" [54, 495]. He was not entirely satisfied with the term "individual" and would drop it whenever he learned a better one. Nevertheless he conceived of each continent as a "natural Whole"--in terms of all its characteristics [55, 518 f.]. Similarly, he wrote, in some cases at least, of the divisions of the continents, die Länder, as "individuals" that are "members,"rather than mere parts of tile "organisms" of the continent [cf. Bürger, II, 14-19]. Although these concepts fit Ritter's teleological philosophy, they are not necessary consequences of that philosophy, nor is such a philosophy necessary to those concepts. In any case we are not to take the terms literally [cf. Hözel, 27].

Humboldt does not appear to have considered this concept. Döring concludes that "whereas for Ritter the individual area had its own particular determined value, for Humboldt it was only a variation of the great cosmic theme of law and causality" [22, 162]. "Nature in every corner of the earth is a reflection of the Whole. The forms of the organism repeat themselves in ever different combinations" [60, II, 89]. In a sense Humboldt was not faced with the question of regional division since he made no attempt to organize the regions of the world, or of any major part, into a system. That he recognized the need for such an organization of geographical study we will note later.

If we put aside the question of the philosophical viewpoints of the two founders of geography as irrelevant except as it affected their work--and in that respect by no means so important as one has been led to believe--what are the major differences between the directions which they gave to geographic work?

Although we find it erroneous to draw a fundamental contrast between them in regard to the proper content of geography, it is nevertheless true that the difference in character of their work led to an important difference in later interpretations. Though Humboldt, at various places in his writings presents brief systematic considerations in anthropological geography [Peschel, 66, I, 351-5], his studies in systematic geography were primarily limited to non-human features. In Ritter, on the other hand, the interest in the study of non-human elements was subordinated to the interest in man, and his great work is presented in the form of a regional geography; further his methodological studies repeatedly emphasized the importance of regional organization of geography in contrast to that by classes of phenomena. In consequence of this double contrast we find the confusion between the two forms of dualism in geography to which we have previously referred-- namely, that the distinction between human and natural (non-human) geography is contained in the distinction between regional and systematic geography.

If we separate these two issues, we find, as already noted, that the inclusion of human as well as nonhuman features in geography did not raise any question of dualism in the thought of either Humboldt or Ritter, but rather was absolutely essential to their fundamental concept of the unity of nature.

The second form of dualism, however, was unquestionable recognized by both men as a contrast in form of study and organization of geographical material. Furthermore, this dualism in methods has presented in issue that has recurred repeatedly in the history of geography, including the present time, and geographers have not been lacking who wished to solve it by the simple method of excluding either one or the other aspect--that is, by limiting geography either completely to systematic studies or completely to regional studies. What was the position of the two founders of modern geography with reference to this issue--as expressed both by their theoretical statements and in their practice?

There can be no question that Ritter's emphasis on regional organization, both in his work and in his methodological statements, tends to give the impression that he had little interest in systematic studies of each analyzed element over the world. But if our comprehension of a student, generally regarded as of major importance in the development of geography, is to consist of more than mere impressions based on the outline of his work and a few sentences extracted from his methodology, we must examine this question more carefully than did the young critic of a century ago [Fröbel, 54].

The present issue is not to be confused with the related question, previously discussed--the argument that Ritter attempted to synthesize the factors associated in an area without preliminary analysis. That question might be dismissed with the mere statement that no matter what he may have appeared to say, in a few sentences removed from their context, the man whom successive generations of eminent scholars regard as a master of geography was no fool. The present question, however, is a genuine one.

Ritter's methodological discussions are to be regarded less as balanced treatments of a field of science than as the program of a reformer attempting to reconstruct a previous geography into a science. In the systematic studies of certain of his predecessors it seemed to him that the intellectual ("subjective") process of dividing the interrelated phenomena of nature into separate classes and studying these separately over the world was a disruption of the actual coherence of nature [49, I, 20]--a conclusion, we may note, that requires no teleological basis. He observed that, when each of these kinds of phenomena is studied in its forms and processes, the result is not only comparable with that of other sciences, but is rather a part of some other science. A "compendium" of these separate studies would not form a geography and the aggregation of areal sections of each of them for any particular area would be no science at all."²¹

Ritter therefore insisted repeatedly that the proper method for a science of the earth was "to ask the earth itself for its laws"; it must use the "objective" (we would say "empirical") method, "calling attention to the main types of formations of nature"; by investigating the relationship grounded in nature itself, this method leads to a "natural system" [49, I, 20; cf. Wisotzki, 1, 273 ff.; Bürger 11, 15].

On the basis of such statements, then, one might, like Fröbel, accuse Ritter of ignoring the importance of systematic studies of individual futures. To be sure, one finds statements which indicate the opposite, but one must look fairly carefully to find these. For example: "the earth's surface its depths and its heights, must be measured, its forms arranged according to their important characters" [49, I, 4]. Further in the lectures given before the Academy of Sciences in Berlin, in 1828 and 1833, one will find considerable discussions of the need for studies of individual categories of phenomena [50, 129 ff., 152 ff.].

Nevertheless, most later critics have felt that Ritter's regional work would have been more effective had it been based more thoroughly on systematic studies, even though the resultant delay would have limited its scope. This criticism was never effectively presented during Ritter's lifetime. It might be inferred in Bucher's remarkable critique of 1827-- to which have already referred--though that is certainly not directed at Ritter. Whereas the Erdkunde is barely mentioned, Ritter's earlier systematic geography of Europe is treated at length as the principal example of the correct systematic approach [51, 93 f. ] But whether Bucher intended to imply a criticism of Ritter's later work, or whether Ritter was conscious of that, I not know; the work was certainly called to his attention by an editorial footnote to his own article in Berghaus Annalen, in 1831 [55, 516]. Neither Marthe nor Wisotzki, both of whom recognized the importance of Buche study, appear to have inferred any criticism of the Erdkunde.

Much better known is the critical essay by Julius Fröbel, at that time, a young student with two years' experience in academic geography (on Fröbel's own program, see Sec. III, A). On Ritter's recommendation Berghaus published the article in his Annalen and the fact that Ritter commented on it at length in the same issue focussed the attention of the academic world upon it [54; 55]. Describing the episode some fifty years later, Fröbel speaks of his surprise and embarrassment at finding himself, a young student of no particular attainments at the time, thrust suddenly into the academic limelight as an opponent of the master. It is an interesting indication of Ritter's character that in his published remarks about Fröbel's critique, he ignored the easy opportunity to make Fröbel look foolish, and on the contrary commended his critical efforts as an expression of an "earnest thought for the furtherance" of geography from an "unmistakably keen mind" [55, 506 f.). Further, when Fröbel came to Berlin shortly after, Rittter received him cordially and the young student received much needed employment working on the maps for the Erdkunde. Fröbel's account of his reception by Humboldt is also interesting, but his statement of the latter's apparently favorable reaction to what he himself termed his "impertinent" criticism (of Ritter) is not specific and has only the degree of credibility granted to personal reminiscences written more than fifty years after the event [28, I, 60 ff.].

In spite of the attention which Fröbel aroused, the net effect on the geographic thought of either Ritter or his contemporaries, appears to have been nil. Whatever degree of truth there may have been in Fröbel 's criticism was lost sight of because his attack, based on but limited knowledge of Ritter's work, so widely overshot its mark--in striking contrast, we may note, to Bucher's critique which had been based on very careful analysis of the works criticized [ cf. Marthe, 23, 29 f.; Fröbel does not mention Bucher's study, not even in his succeeding article, after Berghaus had rather pointedly called attention to it].

Although it was Ritter's "point of view" and "method of treatment" in geography, that Fröbel subjected to criticism, his consideration ignored Ritter's methodological essays and confined itself to the two volumes of the Erdkunde that had been published up to that time. On the basis of a few sentences in the introduction of the first volume, some of which he misunderstood, Fröbel arrived at a critical conclusion concerning the value of the work as a whole. In one case, as previously noted, the criticism is valid only if one considers the single sentences removed from their context. Fröbel 's discussion of Ritter's use of the term "comparative," together with the latter's comments in reply, did little to clarify that question, since, as Plewe notes, each party proceeded from different assumptions and ignored those of the other [8, 38]. Undoubtedly Ritter's lack of precision in choice of words contributed further confusion to a problem that had previously bothered geographers, and over which many later students have wrestled; the most thorough and illuminating discussion of the whole question is presented in Plewe's "Untersuchungen über den Begriff der 'vergleichenden' Erdkunde" [8].

On the issue with which we are here concerned-- the relation of systematic studies to regional geography--Ritter's theoretical statements offered clear grounds for argument. In reaction to those who had specialized on systematic studies and never reached the actual formations of interrelated factors in areas, he had called first for a study of the areas filled with interrelated phenomena, asserting that, when this had been done for the whole world, the material would then be available for a more successful development of the general principles of systematic geography [49, I, 75]. Ritter maintained his belief in this methodological principle in many later writings, notably in his essay of 1833 [50, 181].

In assuming, however, that these theoretical principles constituted the foundation for Ritter's Erdkunde, his critic ignored, or was ignorant of, the fact that elsewhere in his methodological discussions Ritter had shown the need for the opposite method, and that in practice he had in large part followed it. We have already noted his early study of the major geographic features of Europe that Bucher had cited as the model for systematic studies, and the handbook of general (systematic) geography. Although the latter had not been published, various sections had been used in different school texts, in some cases with the original authorship indicated. Ritter could therefore point to these, in his observations on Fröbel's critique, as well to the systematic discussion of individual categories of features that fills more than fifty pages of the introduction to the Erdkunde. Whereas Fröbel had overlooked these, Humboldt and other writers had found in them general concepts and principles that could be used in their own geographic writings--as Ritter observed with detailed references [55, 511]. In subsequent volumes, we may add, Marthe has counted no less than twenty-four systematic studies, some running over a hundred pages each, dealing particularly with minerals and cultivated plants; in many other cases, however Ritter refers to studies to appear in later volumes that never appeared [25, footnote 14]. (See Supplementary Note 7)

Further, the essays read before the Academy of Sciences in 1826 and 1828, include methodological discussions of systematic studies [50, 103-28,129-51]. In the latter, for example, there is a discussion of the study of geometric forms of areas of different categories of phenomena, an early example, we may note, of studies of "patterns." Finally, Ritter could point to his lectures in general (systematic) geography that had been heard by hundreds of his students during the preceding decade. Although he had not published these, they had actually appeared in modified form as Berghaus' Elementen der Erdbeschreibung-- the very work that Fröbel had praised as the model of what should be done in geography! [54, 500, 505]. Berghaus himself, as editor of the Annalen, confirmed this statement in two emphatic footnotes--not without implying that the fact should have been sufficiently clear to any reader of the Elementen [513, 516].²² The irony goes even deeper than Ritter recognized. In repeating the statement more clearly in the introduction to the first volume on Asia, the following year, Ritter claimed that the inner organization of Berghaus' work was the result of his lectures, but gave his friend credit for having improved the style of presentation for elementary students [49, II, xv and 20 f.]. Lüdde however found, by comparing the originals, that Berghaus had published straight dictation of Ritter's course--on many pages, word for word [Plewe, 8, 59].

It is interesting to note that Ritter, who in his youth had been scathingly ridiculed by one reviewer for presuming to criticize his elders--in a work that at the same time offered positive materials--should have limited his personal criticism of Fröbel to the polite but pointed request that he should make himself more familiar with a man's work before attempting to subject it to a critical judgment [55, 520]. It is not surprising, therefore, that Ritter's contemporaries should largely have ignored the degree of truth included in Fröbel's immature and misinformed critique. On the contrary many appear to have followed the more zealously the methodology that Fröbel had attempted to criticize. Undoubtedly this was impressed upon them by the constant emphasis in Ritter's methodological treatments on the full studies of areas, as well as by the contrast between the nineteen volumes of the Erdkunde and the small amount of distinctly systematic studies published under his name. With the notable exception of Reclus, they did little to further the systematic studies that Ritter had made, and in their regional studies they depended even less than he on systematic materials already available. The fact that so much of the work of Ritter's followers later came to be regarded as of little value might be taken as a demonstration of the argument that Fröbel had directed at the Erdkunde: that it could not attain scientific value because it had not been preceded by systematic studies. Indeed, the argument against the Erdkunde, itself, though evidently exaggerated, was perhaps ultimately vindicated. Although scientific travelers, as well as geographers, for many decades--in Germany and especially in Russia--found in Ritter's work "the essential register of facts on Asia" [Marthe, 25, 25], yet for modern students it appears to offer little of importance even to the historical geographer (as distinct from the historian of geography). On the other hand, one may remember, Ritter had found that his earlier attempt to provide first a systematic geography of the world had not been regarded by Buch as successful. Consequently the relative success or failure of Ritter's work does not lead to a clear conclusion as to the proper order of study of systematic and regional geography [cf. Marthe, 25, 23 ff.].

Humboldt's approach was undoubtedly different in many respects. Even less than Ritter, did he regard geography as an end in itself. In order to establish the unity of the total cosmos, it seemed more important--at least for the time--to make systematic studies of particular kinds of phenomena in their interrelations in areas, than to prepare complete studies of individual areas [60, I, 65; 47, I, 2]. Though he made many of both types of studies, his effort to cover the whole of "physical geography" in the Kosmos demonstrated his earlier statement that he regarded this as the more important of his two purposes. In referring to that well-known quotation however, one must understand clearly what he meant by the term "physische Erdbeschreibung"--the expression which he finally adopted after trying various others, such as "physics of the world." We have already noted that in its scope it included all earth phenomena outside the individual student's mind: man and his culture, as well as other organic and inorganic phenomena. Further, it was not the individual categories of phenomena in themselves that primarily interested him, though he had been "passionately devoted to botany and parts of zoology," but rather the interconnections of different categories of phenomena--i.e., of plants of different kinds with each other, and with the differences in climate, relief, and soil and in relation to animals and human life. In addition to the types into which everything that is, may be classified in terms of its intrinsic characteristics, one perceive types in the arrangement, distribution and mutual relations of things of different categories. "It is the great problem of physical geography, (physique de monde) to determine the form of the types, the laws of these relationships, the eternal bonds that enchain the phenomena of life with those of inanimate nature" [47, I, 2-6, quotation on page 6; 44, 2 ff., 33-6]. Throughout all his geographic work Humboldt remained true to the principles he had set for himself in a letter written on the day of his departure for America: "On the interrelation (Zusammenhang) of forces, the influence of inanimate creation on animate animal and plant world, on this harmony, my attention shall always be directed" [quoted by Richthofen, 81, 605-7]. More specifically, in the first volume of the Kosmos, Humbolt described physical geography as the study of "that which exists aready (im Raume) together," it studies phenomena as arranged in areas in their mutual relations to all other phenomena with which they form a natural whole [60, I, 49-72, quotation on page 50; cf. 20, I, 274; 22, 129 ff .] .

In contrast with those who attempt to construct geography after the pattern of other sciences, Humboldt regarded it as having a fundamentally different viewpoint from all the other sciences concerned in the study of nature. The individual natural sciences--what we would call the physical and biological sciences concerned with earth phenomena--"study the forms, construction and processes of individual animals, plants, solid objects or fossils" and seek to arrange these in classes and families "according to their internal analogies." Geography is concerned with these objects as they exist together, related to each other causally in an area (Raum). In contrast with physics or botany, in which "the objects of nature are divided according to kinds of objects," geography regards all the objects as a natural whole, as they stand in areal connection, in part with the earthly body, in part with the universe. Cosmography (Weltbeschreibung--the term is taken from Kant) is "the study of all that has been created, all that is (the things and forces of reality) in the area, as a simultaneously existing Naturganzen." It includes both an astronomical part and a geographical part. Finally, geography is closely related to the third point of view in studying the earth--more closely than to that of the specialized sciences--namely, the historical, which is concerned with the changes in time of all the phenomena [42; 52, 14; 60, I, 49 ff.].

Similar statements of the comparison of geography and history may be found in Ritter's writings, including his first publication of 1804. In particular he notes that as chronology provides the framework into which the multiplicity of historical facts are ordered, the area (Raum) provides the skeleton for geography; both fields are concerned with integrating different kinds of phenomena together, each in its respective frame [24, I, 250 f.; 50, 152 f.]. To what extent he was indebted to Humboldt for these ideas we cannot tell; neither is it clear to what extent each may have depended on Kant, who had long before expressed similar ideas in the lectures on physical geography that were not published until 1801-2. (Humboldt refers to Kant in his lectures of 1827-28, but his own first statement of the concept of geography, on which later versions are based, was published in 1793 [42]. )

Heterogeneity therefore is an inescapable characteristic of geography, as Humboldt understood it. On the other hand, as an eminently practical scientist (by which I do not mean an applied scientist) he recognized that to comprehend a whole consisting of a multitude of things in interrelation, it was necessary to understand first the relations of some of the things to all the others. His studies in systematic geography, therefore, concentrated on particular classes of phenomena--particularly vegetation--in relation to all other phenomena in areas. Whereas "botany proper studies the character, organic formation and relations of plants, in the geography of plants descriptive botany is tied to climatology." The development of plant geography required the combination of means of measuring elevation and temperature with a knowledge of plants, and had been delayed until the methods of the physicist could be combined with those of the botanist. In plant and animal geography however "we speak not of the plants and animals, but of the earth-crust covered with plants and animals" [52, 168-71; cf. also 19, I, 64, 74].

In other words, Humboldt recognized that, though a systematic study in geography must be focussed on one category of objects, it was in no sense to be restricted to the objects of that category but was to consider those objects in their relationship to other geographic phenomena. His physical geography therefore cannot be completely divided into a list of highly specialized fields each separate from the other; rather, in each part, one group of features is the center of attention, but its relations with others carries the work into all the other divisions. Since these relationships are those based on areal position--rather than historical sequence--the fundamental unifying principle is chorological [cf. Döring, 22, 51-9].

This chorological point of view Humboldt consistently maintained in all studies that he entitled geography. Even in a textbook for geography that he and Berghaus projected (for schools in India, at the request of certain English authorities) he refused to insert material that properly belonged in botany or zoology. "A physical geography can concern itself neither with energy and material, nor with the physiology of the organic bodies; all that must be assumed as known" [quoted by Berghaus, 19, III, 94].

Humboldt's contribution to the development of systematic geography is therefore to be found not only in the value of the works which he produced but also in the fact that he first clearly portrayed the distinction between systematic, but chorological, studies in geography and systematic studies in the special sciences. That he was for a time regarded as an important figure in such sciences as botany or geology rested on a misunderstanding of his work, as he himself explained it. The judgment of later specialists in these fields, that Humboldt had done little of major importance in their sciences, was an echo of his own purpose [Peschel, 66, I, 310]. The fact that, rather than attempting systematic studies in the special fields, Humboldt, through the greater part of his work, maintained the chorological point of view, enabled him to found several branches of systematic geography, notably climatology and plant geography [Peschel, 66, I, 316 ff., 328 ff.; Grisebach in 20, III, 232-68; Hettner, 2, 309; 161, 85 f.].

On the other hand it would represent a complete misunderstanding of Humboldt's concept of the field of geography to suppose that he thought it limited to what he called "physical geography," i.e., to chorological studies in systematic geography. Since his own work was based on his field observations--enriched by those of a great host of other writers, both scientists and travelers--he never prepared a regional geography of the world or even of a single continent. But his recognition of the need for such studies, co-ordinate with work in systematic geography, is shown by the outline which he drew up in 1848 for the geography text to be used in India, in which "Specielle Geographie: Länderbeschreibung" forms the second half, co-ordinate with the first, Physikalische Erdbeschreibung [19, III, 55-61].²³ The same view is indicated in his discussion of Varenius' outline of geography [60, I, 60; 52, 14, 26], as well as in his many references to the work of his "great and celebrated friend," Ritter. More significant, however, is the large number of explanatory descriptions of individual regions which he studied in terms of the interrelated totality of their phenomena.

Many of these regional studies of Humboldt are interspersed in the record of his travels [47], others will be found in his Ansichten der Natur [45, note particularly "Das Hochland von Caxamarca," 311-34]; an excellent collection of such regional descriptions has recently been published in a single volume of selections from many different writings by Humboldt [48]. If Humboldt considered these as less important than his systematic studies, he certainly did not regard them as unimportant, as is shown by his careful and detailed discussions of the difficulties involved in such studies and of his method of solving them [45, x f.; 47, I, 137 ff.; and, especially in the Kosmos, 60, II, 1-134, "Anregungsmittel zum Naturstudium"].

For Humboldt, "there is a certain natural physiognomy that belongs exclusively to each climatic zone; every vegetation zone, in addition to its particular advantages, has also its particular (or specific) character." His purpose was to gain "a comprehension of the natural character of the different regions of the world" which "is most intimately related with the history of the human race, and with that of its culture" [43, 11-14; Kosmos, 60, II, 90 ff.; cf. Grisebach in 20, III, 267]. Although Hurnboldt's personal inclinations may have led him to prefer to study areas in which man was of little importance, his interest was limited neither to the non-human features nor to the visible landscape features. On the peninsula of Araya (now in Venezuela) his range of observations included astronomical and atmospheric conditions, plants and animals, the salt marsh and the salt works based upon it, the pearl industry, the physical characteristics of the various races represented, and particularly the changes in the character and customs of the native-born Spanish colonists resulting from the differences in natural conditions as compared with Spain. This situation he contrasted with that found in the highlands, or in the United States, where European settlers had found climatic conditions more similar to those in Europe, and with the character of the ancient Phoenician and Greek colonies [47, I, Book II, Chap. IV].

Similar considerations may be found in what might be regarded as Humboldt's more scientific works. "How powerfully has the sky of Greece influenced its inhabitants! Were not the people who settled in this beautiful and fortunate part of the earth, between the Oxus, the Tigris, and the Aegean Sea, first awakened to moral grace and sensitive feelings? And did not ... our ancestors bring anew gentle manners from those gentle valleys?" "The influence of the physical world on the moral, the mysterious interrelation of the material and the immaterial, gives to the study of nature, when one raises it to a higher point of view, a too-seldom recognized charm of its own" [43, 13 f.]. Peschel appears to have overlooked such statements in Humboldt's writings while objecting to similar suggestions in the work of Ritter's followers, specifically of Kapp. In particular, Peschel attacks the assumption of "a causal relation between the impression of the Landschaft and the mental expressions of the population" [66, I, 404].

On the other hand, Humboldt's descriptions of areas, Peschel found, had introduced a new period in regional geography (Länderbeschreibung). He succeeded not merely in presenting the picture of the area, but also in "making it live through the play of forces;" thus he "made real the interconnection of the observations with a higher order of the Whole." "The post-Humboldtian geographer is concerned not only with the interrelations of physical phenomena, but also with those of the historical [human] phenomena with their scene." In particular, Peschel finds that geography was raised to scientific stature in the regional studies of Cuba and Mexico ["New Spain," 46]. In these studies, to be sure, Humboldt's interest in natural conditions is subordinate to that in social conditions and much of the material, according to Hettner, is not to be included in geography [161, 84]. Humboldt himself apparently regarded these as studies in political economy. Nevertheless the manner in which he described the differences in economic, social, and political conditions of different areas in relation to the differences in natural conditions was in fact followed, not by economists nor political scientists, but by geographers, who thereby, according to Peschel, made this form of regional study one of the functions of geography [66, I, 336-44].

In his study of the development of economics and of economic geography, Schmidt entitles Humboldt's study of New Spain "the first scientific, political and economic geography founded on the nature of the land" [7, 74]. Paradoxically we may note, while Humboldt thus made a major contribution to regional economic geography, Ritter, in his incompleted studies for a "geographische Produktenkunde," was one of the first to make systematic studies in economic geography [Schmidt, 7, 84, referring to the systematic studies in the Erdkunde, previously mentioned here, and to 50, 182-206].

In general, we may recognize two major differences between the two founders of geography in their actual development of regional geography. Ritter's ambition to produce a regional geography of the world--even though his long life did not permit him to carry it beyond Africa and Asia--led him to consider regions of relatively large size in which he seldom if ever completed a full study of all related phenomena.²⁴ In contrast, Humboldt described in great detail areas he had visited, some quite small, and so was able to provide--within the means of observation then available--a relatively complete picture. In order to accomplish this he not only considered each of the significant features of an area but combined these individual characteristics for each distinctive part of the area, then showing the relation of the different parts of the area in the total [47, I, 137 ff.]. Indeed the great attention which he gave to certain small areas would justify the students of "microgeography" in proclaiming Humboldt as their leader.

The second, more important, difference we have already indicated. Ritter maintained in theory, and in part in practice, that geography should first study all the interrelated phenomena to be found in each of all the areas of the world, on the basis of which systematic studies could be made of the relations of individual types of phenomena. The opposite view, that Fröbel proclaimed, called first for a complete study of the individual phenomena systematically over the whole world. Humboldt, as a practical scientist, evidently felt he must do what he could do: he could more easily consider the relations of the two sets of phenomena--e.g., plants and elevation--than those of all phenomena found together; furthermore he could not wait for a complete world regional survey. On the other hand he felt called upon to provide an explanatory description of the lands which he had studied as fully as he could study them. For this purpose he could not wait until systematic geography had been completed. But, by using the results of his systematic studies in his descriptions of areas, he produced "masterpieces" of regional interpretation [Hettner, 161, 86].

The ultimate results of these differences are both interesting and paradoxical. Ritter's relative neglect of systematic studies not only detracted from the value of his regional work, but set a tradition which for a time limited geography in such a manner that regional geography itself could not progress. In the case of Humboldt, on the other hand, the great thesis of the cosmos, that was of first importance to him, is now of little moment either in science or philosophy. Though his studies in systematic geography were of major importance in the historical development of the field, they have long since become obsolete. In part the same is true of his studies in what has come to be called "comparative regional geography" in which he, rather than Ritter, was the pioneer [Hettner, 161, 403; Plewe, 8, 46-55]. But that part of his work which he once indicated as of lesser importance, the masterly explanatory descriptions, analytic and synthetic in form, of individual regions of tropical America and Mexico, remain of imperishable value in geography. As Lehmann noted of any successful studies in regional geography: they cannot become obsolete, they provide irreplacable material for historical geography [113, 239; cf. Sauer, 84, 185].

Brief consideration may be given to certain other effects of the work of Humboldt and Ritter on the subsequent direction of geographic thought. Humboldt was always greatly interested in the aesthetic aspects of geography, "as a means of stimulating and widening scientific nature study." Though he found models of such description in the writings of Goethe, Rousseau, de St. Pierre, and various French and English travelers and literary writers, previously mentioned, it was particularly his "illustrious teacher and friend, George Forster," who had "most effectively pioneered in this direction" [60, II, 74]. Humboldt insisted that this aspect of geography was not to be distinguished from scientific geography; he was evidently unmoved by Fröbel's specific demand that such studies should be grouped with Ritter's historical-geographical studies in a non-scientific "historical-philosophical geography" [56, 7 ff.]. Although he does not mention Fröbel's essay, his statements in the second volume of the Kosmos present a specific reply to the former's suggestions. It was not the description of his own feelings that the author of nature descriptions was to present; on the contrary by describing objectively the external nature surrounding him he should leave complete freedom for the feelings of the reader. The geographer's descriptions of nature, we may conclude, are not to be artistic in the sense of expressionistic; neither are they to be artistic in the sense of form. The aesthetic aspect is not to be supplied by "poetic" phrases of the writer, but by the actual scene itself which it is merely his function to reproduce as well as his command of language may permit. "Descriptions of nature can be sharply limited and scientifically exact without thereby losing the living breath of the power of imagination. The aesthetic aspect must proceed from the presentiment of the interrelation of the sensual with the intellectual, from the feeling of the universality, reciprocal limitation, and unity of nature-life." It is not alone the description of the view naively beheld that is to provide the intellectual pleasure, but far more the understanding of the "harmonious interplay in the forces in the landscape" [60, I, 34; II, 72-4; cf. Döring, 22, 89; Peschel, 66, I, 336 f.; Grisebach, 20, III, 251 ff.; Dove, however, felt that Humboldt had himself been guilty of employing the poetic form of description to which he objected in theory, 21].

Neither Humboldt nor Ritter give a categorical answer to the question of the physical scope of geography--whether it includes the earth as a whole or is limited to the earth surface. No doubt one may find in Ritter's writings, as Gerland was able to do, statements that indicate the former, but it is clear that these are in conflict with the greater part of his methodological discussions as well as with all of his work: the "earth" which he described as the show place of the forces of nature, and particularly as the dwelling place of man, was not the earth body but its enveloping shell. (See Supplementary Note 8 ) It was characteristic of Ritter that he should have been primarily responsible for establishing the term, Erdkunde or earth science--a term, as Hettner observes, that so poorly expressed his own concept. Whereas for Ritter the word was used as an alternative to the word Geographie, of foreign origin, and to the less scientifically sounding Erdbeschreibung which Humboldt favored, later students came to interpret it literally as the science of the earth.

Humboldt, in order to establish his thesis of the unity of the cosmos, took the entire universe in his view, but within this distinguished "Erdbeschreibung" as the "telluric or earthly (irdisch)" part [60, I, 51 f,]. His own research was largely limited to the earth, and indeed, to that part of it which he knew--the earth surface [Döring, 22, 55]. Consequently the tendency of the pre-classical geographers to confine their field to the earth surface was, in practice, continued through the classical period.

In one other respect, Humboldt's actual work extended beyond the field that Ritter considered, namely in the study of the distribution of different types of phenomena--essentially one may say, "the Where of things" [Döring, 22, 64 ff.]. He may, however, have regarded such distributional studies as merely preliminary investigations necessary for the systematic studies of interrelated phenomena; in any case they constitute but a part of his geography.

Finally, we may add, in the development of the methodology of geography during the middle of the nineteenth century--if not longer--two particular factors caused Ritter's influence to be of much greater importance. As the holder of the only university chair of geography in Germany--if not in the world--he exerted a great influence beyond the range of his immediate students, whereas Humboldt's influence was for some time greatest with men who did not consider themselves geographers. Secondly, whereas Ritter repeatedly expounded his views on the nature and problems of geography in methodological papers, Humboldt's numerous discussions of such questions were scattered through his general writings. For this reason, perhaps, his concept of the relation of geography to other sciences was lost sight of for nearly a century.²⁵

The "classical period" in the development of geography may conveniently be considered as terminating with the death of both Humboldt and Ritter in 1859. During the following decade academic geography was dominated by the school which Ritter had founded. Since his followers tended to emphasize the "historical" aspects of the field even more than had Ritter, one may speak of a drift away from systematic geography to a regional geography primarily concerned with man. According to Peschel, Ritter's "Single laudable student" was Ernst Kapp, but Kapp's interest extended also into political problems, and following 1848, disciplinary action led him, like Fröbel, to come to the United States, where he settled as a cotton-farmer in a German community in Texas and apparently did not again engage in geographic studies until after his return to Germany in 1865 [see Peschel's review in 66, I, 399-413, 418; see also Fröbel 28, I, 477 f.; biographical facts from 29]. Another follower, if not student, of Ritter to come to the United States, was the French Swiss, Arnold Guyot, who held perhaps the first chair of geography in this country, at Princeton. That he had but little influence may perhaps be accounted for by the manner in which the teleological view dominated even his interpretation of details [64].²⁶

By far the most successful of Ritter's students and followers was the French geographer, Elisée Reclus. According to Girardin and Brunhes, it was his study under Ritter, in 1851, that made Reclus a geographer; from Ritter he derived his main principles and ideas concerning geography [30, 67-9, 71]. It is interesting to note that Reclus also followed Ritter's practice, rather than his theory, in first preparing limited studies in systematic geography and then a general systematic geography, before proceeding to a complete regional survey of the world. Unlike Ritter, however, he established his reputation by his systematic study, a major work in physical geography (La Terre, 1866-67; later translated into English). Although Girardin and Brunhes found that the influence of Ritter especially predominated in this work, other students have noted the dependence on many other writers, including the English student, Mary Sommerville. (See Supplementary Note 9) Spörer praised this work as superior to any of its kind in the German literature, contrasting it with the work of the Ritterian school, apparently not recognizing Reclus as a student and follower of Ritter [68, 331 f.; presumably the same is true with regard to Peschel's statement quoted above]. Hettner however finds the work distinctly limited by Ritter's influence [161, 108]. Schmidt's statement that Reclus "became the Ritter of France" no doubt is based largely on his nineteen-volume regional survey, Nouvelle géographie universelle, 1875-94. Though this obviously followed Ritter's Erdkunde, it was far more successful, "both in the taut organization, which permitted him to complete the great work, and in the close coherence of nature and culture in every earth area" [7, 151-3].

Ritter's influence was also significant in certain other fields. One of his most interested students at the military college was Moltke, who published various geographic studies long before his military-geographic plans became of momentous importance [Schmidt, 7, 86]. Likewise Ritter was influential in calling the attention of historians to the significant relation of geography to the course of human events; Marthe notes this influence particularly in the work of the celebrated historian E. Curtius, a former student of Ritter's [25, footnote 10].

Though Humboldt, who held no university post, had no immediate followers in academic ranks, his influence outside the universities--and outside of Germany--was vastly greater than that of Ritter. During the brief "post-Ritterian period" Hettner finds that "the real representatives of the true geographical science were the scientific travelers who took Humboldt for their model" [2, 313].

The post-Ritterian period however proved to be but a brief interlude before the last quarter or third of the century brought a very rapid development in academic geography in Germany. In many respects this period may be regarded as the critical period in the development of the field. The foundations which Humboldt and Ritter had established for geography did not provide, in appearance certainly, a clearly unified field. To the extent to which their followers exaggerated certain aspects of the views of each of the founders, or attempted to introduce new concepts of the nature of the field, geography was for a time split in several directions and its position as a branch of knowledge thereby brought into serious question. Following the death of Ritter there was no professor of geography in any German university and the return to university status and particularly the rapid subsequent growth was largely the work not of the "historical geographers" who followed Ritter, but of students who had been trained as geologists and tended to specialize in the study of non-human features of the earth--i.e., physical geography as we understand the term [cf. Penck, 129, 635 f.]. With the rise in academic status of geography and the productive work of this period we are not here concerned; in the development of geographic thought, its major problem was to overcome the apparent disunity in the methodology of the field and thus definitely to establish its position as a single field of science.

The general scientific atmosphere during the latter part of the nineteenth century was far from receptive to the philosophical concepts of the earlier "romantic" period, whether those of Ritter or of Humboldt. This change was evident even before the end of the classical period. Indeed the differences in geographic thought overlap our arbitrary divisions of its historical development to such an extent that we find certain of the concepts of the post-classical period expressed by the last of the pre-classical geographers, Bucher. We have already noted Bucher's thorough attack on the use of "natural boundaries" and "natural regions" in the works of his contemporaries. Utilizing anatomy and physiology as analogies--though recognizing that the analogies were not complete--he arrived at the negative conclusion that geographers need not attempt in any way to divide the earth into areal parts except for special purposes; rather that they should study it in terms of classified phenomena, i.e., systematic geography [51, 90-94]. Although Bucher used Ritter's study of 1806 as an example of what should be done, and speaks in other connections in praise of Ritter's work, his criticism unquestionably was applicable to the volumes of the Erdkunde, whether he realized that or not.

Though Berghaus appears to have been influenced by Bucher's study, and Marthe was evidently familiar with it, it is not clear that it had any important significance in the development of geographic thought, possibly because of its publication by a small provincial press [cf. Spörer's discussion, in a similar connection, 68, 365]. On the other hand the very similar views which Fröbel presented a few years later in the principal geographic journal of the time created quite a stir in academic quarters, particularly because they were directed at Ritter and were accompanied by Ritter's comments.

The controversy however proved to be a flash in the pan; Fröbel's demands "died away without effect" and Hettner therefore omits them from his consideration of the development of geographic thought [2, 305], That view may be somewhat exaggerated however, since Fröbel's articles, at least the two published in Berghaus' Annals, retained a fairly conspicuous place in geographic literature and may well have influenced later students whether they realized it or not (see also Sec. III A). Certainly the viewpoint which he was trying to express ultimately came to the fore and Plewe therefore regards him as a herald of the new period in geographic thought-- preceded, we may add, by Bucher [51, 59 f.].

The new viewpoint in science which came to dominate scientific thought, Plewe characterizes as marked by the increasing specialization of the sciences, by an increasing emphasis on the development of "scientific laws," and by a conscious isolation of science--and specifically of geography--from any particular Weltanschauung: this latter point of view, however, often represented an equally definite philosophical presumption of a materialistic, inechanistic, universe in which man was to be studied as a "thing," like any other--"a sum of atomic movements" [Plewe, 8, 60, 66 f.].

The shift in geographic work is generally regarded by German geographers to have been due primarily to the work of Peschel and Richthofen. The movement was originated in the essays which Peschel published from 1866 on, a number of which were collected in 1870, in "Neue Probleme der vergleichenden Erdkunde als Versuch einer Morphologie der Erdoberfläche" [67]. "It has been quite justly said," writes the Belgian geographer Michotte, "that with this work the scientific spirit re-entered geography" [189, 24]. Through this work, and through his teaching, Peschel led geographers to study primarily the morphology of landforms. His geographical work was by no means limited to this part of the field. On the contrary he was concerned also with the study of the influence of landforms on human history, and was successful, Marthe found, in limiting this more narrowly and sharply than Ritter had done [25, 22]. His early death, at the age of forty-nine, cut his work short just at the time when his university position should have enabled him to be most productive. (For a more adequate treatment of Peschel's contribution--as well as a biographical sketch--see the study by Ratzel [31], as well as briefer statements by Schmidt [7, 147-9], and Döring [22, 163-51]).

The scientific morphology for which Peschel had striven was given its foundations by geographers who had been trained as geologists--notably by Richthofen, in his studies of China, published from 1877 on (Plewe, 8, 74; Hettner, 161, 99, and especially 32; Penck, 128, 43-51, and 137; and Schmidt, 7, 153-6]. Among the followers of Richthofen in this particular direction we may mention two: Penck--also trained as a geologist--whose classical study of "Die Morphologie der Erdoberfläche" first appeared in 1894, and whose many students are still active in this field; and the American, W. M. Davis, part of whose work, as we have noted, is included in the German literature.

In consequence there was inaugurated in Germany--as later under Davis in America-a long period when geomorphology represented the major field of geography [Philippson, 143, 9 f.]. The inclusion of this subject within geography appears thereby to have been irrevocably established in Germany, whatever logical objections might be raised [Kraft, 166, 7], whereas in America, its position with reference to geography and geology is still uncertain. (For a further discussion of this question see Sec. XI G.)

Whatever views might be held in regard to geomorphology, it has been commonly recognized that under Peschel's leadership, geography for a time tended to expand in the natural sciences into fields that had long been claimed and cultivated by other sciences, so that it seemed for a time as though geography would claim all of physical science that was related to the earth [Hettner, 2, 314 f.]. Human geography on the other hand--including ethnography, agricultural land use, trade, and the movements of peoples, as in Richthofen's studies of California and of China [69]--was considered largely in relation to landforms, or was confined to studies in regional geography. In the development of regional studies particular mention may be made of Kirchhoff, as a connecting link from Ritter to Ratzel [33; 34].

One important result of this situation was the confusion, to which we have previously referred, of two forms of dualism in geography into one: a systematic, physical (now meaning non-human) geography, and a regional, human geography. In order to justify the inclusion of studies of man in geography, and at the same time to prevent the indefinite expansion of geography in that direction it seemed necessary to state the purpose of the two parts separately. Natural features were studied in their own right--presumably it mattered little how far the field expanded in that direction--but human features were studied only in terms of their relations to the natural features. This view was more or less definitely stated by Wagner, Kirchhoff, and Neumann [according to Hettner, 2, 316].

The outspoken emphasis on the division of the field into two sets of phenomena pointed logically to two opposite directions of development. That which came to have the more important effects can be traced from Ratzel through Semple to America where it had its logical ultimate expression in Barrow's statement of geography as the field of "human ecology," the mutual relations between man and the natural environment [208]. This view has been echoed in England [195], and also in Japan [110]. Because of its great importance in this country we shall need to consider it later in detail.

The opposite conclusion--that man should be excluded entirely from geography--was urged much earlier, namely by Gerland in a long essay in 1887 [76]. Although this statement came "at a time that was certainly open to the ways of thought of natural science" [Bürger, 11, 24], and might well have seemed the logical conclusion of the swing in geographic thought of the previous twenty years, it met with essentially no favorable response. It was clear to many, as Hermann Wagner observed in his lengthy critique, that if Gerland's apparently logical thesis were carried through to its necessary conclusions, so much would have to be excluded from geography that the remainder would not be recognizable as the field that had long carried that name [77]. Few thought it necessary to give much attention to a thesis that would have thrown overboard the greater part of the geography of Humboldt as well as that of Ritter [cf. Supan, 78, 153 ; Hettner, 2, 315 f. ]. Indeed it would be difficult to point with certainty to any effect of the controversy which Gerland introduced on the subsequent development of geographic thought in Germany; hence, while we will have occasion to return to it in other connections (Sec. III A, B), we need not examine its claims here.

The historical significance of Gerland's proposal was suggested a few years later by Wagner: the announcement of geography as a pure natural science came at the high point in the swing toward more intensive cultivation of physical geography that Peschel had started nearly two decades before [80, 374-5]. Although Gerland, as Schmidt remarks, was "simply consistent" with the course of this movement [7, 156], none of the "morphologists" was willing to continue to the logical extreme. Indeed, it is possible that his arguments may have made some realize how far geography had been swung out of balance.

The counter-movement however had already set in, both in actual geographic work and in methodological considerations. A major event in the history of geography that was to have indirect influence on its methodological development was the publication, in 1882, of Ratzel's "Anthropogeographie oder Grundzüge der Anwendung der Erdkunde auf die Geschichte" [72]. The term, anthropogeography, which has ever since been associated with Ratzel's name, reflects his background in the natural sciences, particularly zoology, but is misleading. It suggests the geography of man in terms of individuals and races, anthropological geography; whereas the major objects of Ratzel's concern were the works of man, particularly--thanks perhaps to Moritz Wagner--the products of man's social life in relation to the earth [Schmidt, 7, 157-61].

Ratzel was therefore, like Ritter, a student trained primarily in the natural sciences for whom geography offered the connection between the natural sciences and the study of man. As we found in the case of Ritter, the fact that Ratzel's work was largely in human geography has led many to overlook his earlier background. Thus, Sauer's statement that he "got into geography through newspaper work" [84, 166] is less than half the story. According to Ratzel's own account it was his interest in nature that determined his career; he never departed from his early intention of "devoting himself to some kind of scientific study." He took his doctorate in zoology, geology, and comparative anatomy. Continuing his zoological research in southern France he sought to replenish his funds by sending travel letters to the Kölnische Zeitung, with the result that he was offered a position as scientific and travel reporter for the paper. Karl Andree advised him to use this opportunity to become a geographer, and though Ratzel at first planned to travel as zoologist he did ultimately, after six years of travel in Europe, the United States, and Mexico, make the shift. His first publications however were in zoology, and his work in geography included physical studies of the snow limit in the Alps [35].

Ratzel's purpose was to establish the study of human geography on a scientific basis. Far more successfully than Ritter, Schmidt concludes, he "remained true to the direction of the natural sciences." His great contribution was "to have brought this part of the geography of man, cultural geography, into a scientific system by organization of the phenomena, and establishment of concepts and significant connections of the results obtained" [7, 158]. Although he is commonly considered as primarily a follower of Ritter, in a sense his Anthropogeographie represents the first major systematic study of the geography of man as suggested in the culminating section of Humboldt's "physical geography." The fire of criticism that has been directed against his work ever since, does not lessen its importance in having demonstrated that the human, as well as non-human, aspects of geography could be subjected to systematic study, thereby leading to more reliable interpretations in regional geography. (For a full appreciation of Ratzel's work, see particularly Hassert's study [36].) (See Supplementary Note 11)

In one major respect Ratzel's approach was different from that of Humboldt. Working purposely from the point of view of natural science, he organized his first Anthropogeographie largely in terms of the natural conditions of the earth, which he studied in their relations to human culture. Though this procedure was common among geographers concerned primarily with physical geography, others, such as Kirchhoff, had studied human geography by the reverse method--by considering human conditions in relation to natural conditions. In his second volume, Ratzel himself largely reversed the process, but many of his followers, notably Semple (The reference to Semple is in error; her work demonstrates both of these methods [204]) in this country, maintained his earlier orientation and thereby established a procedure that dominated human geography for some decades--at least in this country. German geographers, however, changed their procedure much earlier, thanks particularly to the efforts of Hettner and Schlüter [130; 131]. In the meantime, however, the fact that human geography was studied in terms of the influences of the natural environment on man led naturally to the concept of the field of geography as essentially the study of such relationships--a concept commonly, but probably erroneously, ascribed to Ratzel himself. (In particular it is misleading to place Schlüter in opposition to "von Richthofen Ratzel, and their contemporaries" as Dickinson does [202, 2; cf. Schlüter, 131, 507 f. ].)

In the year following the appearance of the first major study in systematic human geography, Richthofen presented, in his inaugural address at Leipzig, what came to be regarded as the programmatical statement of modern German geography [73]. His earlier statement, in the concluding pages of the first volume on China, 1877 [69, I, 729-32], had been an attempt to restate the concept of geography that was common with Ritter and Humboldt, but, as Hettner notes, it placed bounds on geography that were broken in the book itself [126, 560]. In the same year, however, Marthe had clearly restored the areal or chorological principle as the dominant criteria of geography, using the terms chorography and chorology which various writers before him, including Peschel, had taken from the early Greek geographers. At the same time, he emphasized the study of distributions, which had formed a part of Humboldt's geography, and expressed geography in simple terms as the study of the Where of things [70, especially 426-9]. Marthe's concepts also found expression in a statement on the nature of geography--drawn up by a committee of which he was a member--adopted by the International Geographical Congress at Venice in 1882 [71, 679].

In his epoch-making address at Leipzig, then, Richthofen took over Marthe's chorological conception and made it a fundamental basis of his concept of geography [Hettner, 126, 552]. In this address, writes Plewe, Richthofen showed himself to be "the actual one to inherit and carry forward the ideas of Humboldt and Ritter, and also, in a consistent sense, to fulfill Peschel. In contrast with the latter, he had the sound, unprejudiced historical sense to fit himself into the course of development and to determine his position exactly" [8, 73; see also Penck, 128, 43-51].

Richthofen had been trained as a geologist, and as a geographer he was primarily interested in geomorphology. Upon his election to the Academy of Sciences, he chose to enter the "physical-mathematical class" in contrast to "the historian Ritter" who had belonged to the "historical-philosophical class" [81, 605]. Nevertheless, he recognized the value of Ritter's work in the development of geographic thought. If one reads his address against the background of the concepts of Humboldt and Ritter one observes how much less he was interested in the contrast between them than in the common elements of their concepts and work. He was able therefore, "following the precedent of Humboldt, to restore the close connection of geography to the natural sciences," and at the same time to restore Ritter's program to its place in geography [Schmidt, 7, 153-6; cf. Döring, 22, 165 f. ; Hettner, 126, 552 f., and 32].

We need not concern ourselves with Richthofen's specific statements; as Hettner later noted, though he set the direction of geographic thought for the future, he had not been successful in finding a sharp formulation for his concept of the field. But the essential thought is clear from his discussion: geography studies the differences of phenomena causally related in different parts of the earth surface [161, 106 f.; 73, 25 ff.]. Likewise important, in the development of geographic thought, was Richthofen's exposition of the relation of systematic and regional geography to each other and to the field as a whole. The actual purpose of systematic geography is to lead to an understanding of the causal relations of phenomena in areas [42 f.], an understanding which may be expressed in principles that can be applied in the interpretation of individual regions, i.e., chorology. (Richthofen distinguished between a first step, chorography, which is non-explanatory description, providing material for systematic geography, and chorology, a final step, the explanatory study of regions, based on systematic geography; but this separation has not been followed.)

The major distinction in geography, between systematic and regional geography (chorology), is therefore not a difference in materials studied--Richthofen recognized that it is not possible to limit the materials to be studied in geography--but a difference in method of study. Because of the heterogeneous character of these materials, they must be studied by classes, in systematic geography. On this basis he recognizes not two but three major groups: physical phenomena, biological phenomena, and human phenomena. Though man is of course a biological object, and to a certain degree may therefore be included with the studies of other animals, the fact that his relations to the earth surface are governed by a host of factors not significant for other animals, requires an entirely different form of consideration. In the chorological studies, the causal relations between all the groups of phenomena unites them in a single unitary study. Not bound by any limited concept of science, Richthofen, like Humboldt, saw no objection to a single science considering different kinds of things that exist together and are bound together by causal connections.

Richthofen's program for geography and the fuller exposition of it which Hettner later contributed (first in 1895, but most fully in 1905) prepared the way for studies in regional geography interpreted in terms of the fruits of systematic geography. This was not new in geography, but rather a return to the method of Humboldt (2, 309 or 161, 86]. Though neglected, first by the followers of Ritter who studied regional geography with relatively little consideration of systematic geography, and later by the followers of Peschel who considered chiefly systematic geography, it had never been lost sight of. Supan however felt that Richthofen's address, in giving more detailed attention to systematic geography, had subordinated regional geography, and, in 1889, urged its more frequent cultivation [78].²⁷ Likewise Hettner vigorously encouraged the development of regional studies--by example, by his methodological statements, and through publication in the Geographische Zeitschrift which he founded in 1895. Nevertheless this aspect of geography hardly reached a position comparable with systematic, notably morphological, studies, until relatively recent times, particularly after the World War led to a focussing of German interests on the full character of European areas, notably those of Germany. The fact that Hettner is so often referred to as a promoter of regional geography should not lead one to suppose that he discouraged the development of systematic geography. Throughout the successive shifts in emphasis of the last forty-odd years he has consistently maintained the necessity for geographic research from both of these points of view. In his first brief statement on the nature of geography, with which he introduced his Geographische Zeitschrift, in 1895, he commented on the neglect of systematic studies by the followers of Ritter, and credited the leadership of Peschel with having restored systematic physical studies in geography [121, 2; cf.2, 310]. His first detailed methodological study, of 1903, was limited to the fundamental concepts and principles of systematic physical geography [123] ; in that of 1907 he considered systematic human geography [130], and in many others, since 1905, he has stressed the importance of systematic, or general, geography, as a co-ordinate part of the field [especially 140]. Finally, in more recent years, he has him self presented detailed systematic treatises covering different categories of geographic features, first on surface forms [361], then on climate [362], and finally, a four-volume work covering the whole of systematic geography [363]. Human geography however is limited to less than a hundred pages of the fourth volume, since Hettner planned to devote a separate volume to this part of the field. Geographers of all countries will hope that nothing will prevent the publication of this volume for which the manuscript is essentially completed. (See Supplementary Note 12)

In his first discussions of the relation of these two types of geographic work, Hettner introduced a somewhat unusual terminology designed to emphasize that there was no sharp separation between them. In a regional study of any extensive area it is necessary to study the notable variations in the individual geographic features systematically. On the other hand, the systematic study of a particular category of geographic features is not made with reference solely to that category, but rather in terms of its chorological relations to one or more other features--i.e., their relations as each varies in different areas. He wished to emphasize, that is, the distinction that Humboldt had clearly drawn, but which nevertheless had often been lost sight of, between studies in systematic geography and those of the special systematic sciences studying the same objects [see especially 60, I, 48 ff.; on the relation of Hettner's views of geography to those of Humboldt, see Döring, 22, 166-8].

In the German literature of the nineteenth century the terms introduced by Varenius, "general" and "special geography" had in considerable part been replace--in the desire to eliminate words of foreign, origin-- "allgemeine Erdkunde" and "Länderkunde." Not only did these suggest too great a cleavage between the two types of geographic study but the term "general science of the earth" expressed a very different concept from that commonly associated with the word "geography." It was forgotten, as Penck has recently observed, that "Erdkunde" was "simply the germanization of the word geography. One took it literally and derived therefrom the task of studying the earth as a whole." Gerland, under whom Hettner had studied, had attempted to follow this reasoning to its ultimate conclusion and even Richthofen was apparently confused by it [90, I, 39]. One consequence, Hettner felt, was that geographers were influenced to make systematic studies of the earth as a whole in which the chorological point of view was lost [126, 559]. Geography became even more dualistic in nature than Wagner had recognized; studies in systematic geography did not produce general concepts and principles concerning the interrelation of phenomena that were needed for regional geographic studies. To overcome this separation of the field into two entirely distinct types of work, Hettner regarded it as necessary to make the chorological concept dominate--as in Humboldt's work--in systematic as well as in regional studies. To emphasize this viewpoint he placed at the head of his outline of systematic or general geography--drawn up in 1889, two years after the publication of Gerland's proposal--the title "Vergleichende Länderkunde." The relation of this form of geographic study to Länderkunde proper, or regional geography, is clearly explained in his studies of 1895 and 1898. Although Hettner dropped this term in the methodological studies in which he most fully developed the relation of the two kinds of study [161, 398-404], later, when he ultimately fulfilled the outline written more than forty years before, he maintained the original title.²⁸

Before leaving this period of the late nineteenth century, we must note certain characteristics of the geography of the classical and pre-classical periods that, for a time at least, almost dropped out of sight. One of these is the concept of unity or Ganzheit in geography. In his survey of the development of this concept in the history of geography, Bürger is able to present little acceptable evidence that it was important in German geography during this period. Relatively few German students appeared to have been concerned with Ritter's concept of the earth as an organic Whole, though Ratzel may have echoed this thought in certain connections, and it was a favorite concept of the great French geographer, Vidal de la Blache [184, 5]. Gerland's effective argument against this concept reflects an increasing demand in science for direct accurate description in place of either mystical assumptions or misleading analogies. The Earth, he concluded, was not an "organism" but simply a "complex of cosmic matter" [76, vi].

To be sure, German geographers of the late nineteenth century continued to study the mutual relations between different phenomena in areas; Richthofen saw the unity of the field as resulting from the causal interconnections of different kinds of phenomena [73, 16 f., 67], and Hettner, as well as others, emphasized the importance in regional geography of presenting not merely some of the characteristics of an area, but the total character, as determined by the interrelated combination of all significant features [161, 217]. But this total character is not thought of as a "unity," an "Einheit," or "Ganzheit" in the sense of the philosophy of early writers, including Humboldt and Ritter.

Likewise, one does not find during this period the concept of the individual region as a unit in itself, a "whole," or an "organism." Both of these aspects of Unity and Wholeness have been returned to geography in recent decades, chiefly since the World War. Its belated development makes it questionable to consider it as a result of the "return to Ritter," of which both Richthofen and Hettner spoke in a very limited sense.²⁹ Certainly neither Richthofen nor Hettner ever supported the concepts; the latter, as we shall see, has frequently challenged them.³⁰ He himself suggests that the concept might be traced from Ritter's teleological concepts [161, 306], presumably through Ratzel [126, 557; see also Plewe, 8, 72, and Bürger, 11, 76]. It was evidently from Ratzel that Vidal de la Blache took his concept of terrestrial unity, of the "earth organism"; though Vidal erred in supposing it was new in geography [184, 5]. But it is also possible that the notable development of these concepts in recent years is not so much a product of the ideas of Ritter and Humboldt, but rather a repetition of a similar cultural phenomenon--the introduction into geography of general philosophical concepts of a particular time and country. In the post-War atmosphere of Germany (I.e., the atmosphere of the 1930's.), "Unity" and "Totality" are powerful concepts.³¹

In the scientific temper of the late nineteenth century there was apparently little place for the consideration of the aesthetic character of landscapes that Humboldt had so notably developed. Bürger cites as sole exceptions of note, studies by Ratzel, Oppel, and Wimmer [74]. Both of the latter continued Humboldt's usage of the term "Landschaft" in the sense of the visual scene.

In sum, we may say that in the latter half of the nineteenth century, under the influence of the development of the specialized natural sciences, geography for a time appeared to be changing into a field quite different in character from that which Humboldt and Ritter had inherited and developed. The emphasis on systematic studies appeared to divide geography into two halves, one a natural science, the other a social science, united only in a study of regions that hardly appeared to be a science at all, in the sense in which the time conceived that term. By the end of the period, however, reaction had set in, so that the direction of geography was again essentially that which it had been before. At the turn of the century, the purposes of geography corresponded in major degree, as Richthofen stated in 1903, "in content and methodology, to the conception which Humboldt had given it" [3, 673, 689]. Geography then, as throughout most of its history, was concerned, as Hettner put it, to study the areas of the earth (Erdräume) according to their causally related differences--the science of areal differentiation of the earth surface [Hettner, 2, 320].

Although Hettner found in 1905 that German geographers were generally coming to accept Richthofen's statement of the field of geography, he, as well as others, felt that Richthofen "had not been entirely successful in finding the sharp methodological expression for his concept"; that he had not followed it through consistently, and, later, in his address of 1903, had somewhat obscured the issue [126, 552-3, 560; 161, 106]. Hettner therefore set himself the task of providing a sound methodological exposition, not of some concept of geography which he deduced in his own mind, but of the concept which seemed most closely to represent the field of geography as it had developed hitherto. He was therefore not merely standing on the shoulders of Richthofen, but likewise, both indirectly and directly, on those of Ritter and Humboldt and their predecessors of the pre-classical period. His dependence on Humboldt can be noted in countless passages; he himself has indicated it in many [cf. 161, 85 ff.]. Döring concluded that in Hettner's synthesis of the geographic thought of the two founders of modern geography, Humboldt stands not beside but before Ritter [22, 163].

Throughout the long series of methodological discussions which Hettner has published from 1905 to the present time--with further studies, one may add, ready for publication--he has consistently endeavored to express the concept of geography as the outgrowth of its historical development. In that sense they represent the culmination of our study of the historical development of geographic thought. Since his purpose in these studies, however, has been to provide the logical basis for the concept historically evolved, we will examine his views in the following section.

Our historical survey may be briefly concluded. The viewpoint which Richthofen had stated in 1883, as interpreted by Penck, Schlüter, and above all by Hettner, came to be widely accepted, so that German geography in the first part of the twentieth century has been marked by a greater degree of unity of fundamental concepts than ever before. In particular, Hettner's methodological discussions have come to be regarded as "classics" in geography which no German scholar would ignore in any consideration of the methodology of our field. Sölch wrote in 1924 that "almost all the scientific geographers of Germany have arrived at similar conceptions, thanks not least to Hettner's work" [237, 56; cf. also Braun, 155, 6-8; and the statement at the Heidelberger Tagung of representatives of 21 geography departments of university rank in Germany, 138].

Likewise the methodological studies of Hettner and others of his contemporaries have had a notable influence on the concepts of geographers of other countries. Among the geographers of continental European countries outside of Germany who have expressed similar views on the nature of geography we may list the following: Berg in Russia [97, 103] and his student Marcus in Estonia [191, 12 ff.]; Granö in Finland [270, 296] ; Arstal in Norway (according to Braun) and Helge Nelson in Sweden (as quoted by DeGeer, who on the other hand expresses a concept similar to that of Marthe [190, 10]); Michotte in Belgium [189]; and in Italy, Giannitrapani and Marinelli (according to Sölch) and Almagià [188]. Both Chisholm [192] and Herbertson (see Sölch) attempted to introduce Hettner's views in Great Britain. In Japan, a number of geographers today follow Hettner, according to Inouye--including Komaki and Watanuki and, evidently, the reporter himself [110, 287 f.]. The chorographical concept was first introduced into this country by Fenneman [206] and is to be found in one of the later discussions of W. M. Davis, in marked contrast to his earlier, better known statement [102, 209 f.; cf. 203]; further it may be discerned in the work of many others, notably of Wellington Jones. Nevertheless it was largely overlooked by American geographers until expounded with great effectiveness by Sauer a little over a decade ago [211; 84]; since then it has become the general concept of the field for a large number, if not the majority, of the active research workers in the United States.

On the other hand, in countries outside of Germany, there is no such general agreement among geographers as to the general nature and scope of their field of study as has characterized German geography of the past generation. In most countries indeed there is marked divergence and conflict of opinion. It is significant to note therefore that geographers outside of Germany have shown but little interest in studying the methodology of their field, as Stamp (Josiah C. Stamp, the economist.) observes in a recent survey [200]--though it should not be supposed that they do not frequently talk about it and publish their individual views.

French geography presents an exception, in the relatively high degree of unity that was established by the predominating influence of Vidal de la Blache. But important as the influence of Vidal, Brunhes, and Dermangeon has been in the development of regional geography, it has resulted from their works, and those of their students, rather than from studies of the nature of geography [cf. Sauer, 84, 171, 180 f.]. Indeed if one read only what these writers say about geography, as distinct from what they have done in geography, one would have to place them simply as modifiers of Ratzel. The most thorough, and in many respects a highly valuable, analysis of geography from this point of view is Vallaux, Les sciences géographiques, which will be referred to frequently in later sections of this paper [186]. It must be expressly stated, however, that no adequate attempt has been made here to survey the current thought in French geography; the reader will find some references in the recent survey of work in that country by Musset [93].³²(See Supplementary Note 13)

British geographers have shown perhaps the least interest in attempting to determine the nature and scope of their field. Chisholm's inaugural lecture of 1908, based largely on Hettner, appears to have had little influence. Both Sölch and Huender (of Utrecht) have noted that different English geographers express widely differing views, though nearly all are subject to the "environmentalist conception" [98; 99; 100]. At least one English geographer, Dickinson, has expressed distress at the lack of coherence in geographic work in England [101]. Roxby's presidential address in 1930 examined the concepts of earlier German geographers and the interpretation of them by Vidal and Brunhes, to arrive at a statement very similar to that of Barrows; but German geographers since Ratzel are confined to a footnote [195, 282 f.]. The following year Mackinder sought to state the unity of geography in terms of natural regions based primarily on the hydrosphere rather than on the lithosphere--a conception that no doubt owes much to Herbertson [196]. Bryan's book on "Cultural Landscapes" shows no acquaintance with the developments in Germany since Ratzel [280; cf. Dickinson's comment, 101]. On the other hand the work of the Passarge school is discussed at length in Unstead's study of regional geography [309] and Stamp's (Josiah C. Stamp, the economist.) survey of economic geography considers American, French, and German points of view [200]. In the last two or three years both Crowe and Dickinson have given critical attention to more recent German views-- in part, apparently, as a result of the stimulus from this country [201; 202].³³ (See Supplementary Note 14)

That American geographers have been more given to discussing and arguing over the nature of their field than to studying the problem is well-known. Serious scholarly studies of this problem, including a due consideration of the findings of previous students, are limited chiefly to two or three presidential addresses, already referred to, and the well-known discussions of Carl Sauer. Insofar as Sauer's methodology is derived from German writers it depends largely on Schlüter. Though he recognizes Hettner's methodological studies as "at their best perhaps the most valuable appraisals of what geographers are trying to do" [84, 182] he appears to have otherwise largely ignored them, perhaps because of a misunderstanding of Hettner that he appears to have taken over from Schlüter (see footnote 48).

Though the majority of American geographers may not have written on this subject they nevertheless have pronounced, but divergent, convictions as to the proper definition and scope of geography [cf. Parkins's survey of "The Geography of American Geographers," 105]. Categorical criticism of specific papers on the grounds that they are "not geography" are common, and highly debatable views of the nature of the subject are published, of all places, in college text-books [cf. Crowe's comments, 201, 10 f.]. One feels justified in asking these critics to give us the fuller benefit of their judgment in thorough scholarly studies of the nature of geography.

The geographers in Germany and the large number in other countries who agree with them in considering geography as a chorological science concerned with studying the areal differentiation of the world, do not suppose that they present a new concept of an old field. On the contrary, as many, including Hettner and Sauer, have noted, this concept may be derived from the work of the earliest geographers, from Herodotus and Strabo [161, 122; 211, 25]. Modern geography, Sauer f the most ancient geography"; it is appropriate therefore to use the term common in ancient geography (and revived by Marthe and Richthofen), namely, chorology, the science of regions.