Paradigm, No. 23 (July, 1997)

The School Atlas, 1885&endash;1915

 Michael Wise

Department of Geography,
London School of Economics and Political Science,
Houghton St.,
London WC2A 2AE
45 Oakleigh Avenue,
London N20 9JE.


‘Featureless atlases and wall maps, the value of which was estimated mainly by the number of names which they contained.’ So ran John Scott Keltie’s opinion of the quality of school atlases in Britain at the time of his enquiry into the state of geographical education in 1885.1 Keltie’s Report was a key event in the crusade by the Royal Geographical Society to promote geography as a university study and so educate good teachers who would transform the teaching of the subject in public schools.2

The progress that had already been made in the ‘lower’ schools has generally been overlooked, though Keltie himself remarked upon the achievements of the better School Boards, including London, Birmingham, Edinburgh and Glasgow.3 Good work was also in progress in many teacher training colleges. For that level George Philip and Son offered their Training College Atlas which had originally been designed and the maps drawn by William Hughes, though by the time of the 1880 edition the work of E. G. Ravenstein was also included. That edition was priced at 18s. and one wonders how many students were able to afford to buy their own copies! There was much less progress to report in the Public Schools. In 1868, at the instigation of Francis Galton,4 the Society had decided to award Medals for good work in Public Schools but these had had only a limited effect: Dulwich College, where the excellent James Robinson was in charge of geography, and Liverpool College, where Galton’s brother-in-law the Rev. Dr. George Butler was Principal, had carried off most of the Medals.

The educational publishers were alive to the new situation. The impetus given by the Education Act of 1870 had stimulated the production of new maps, atlases, text-books and appliances.5 The Bartholomew family of Edinburgh had introduced layer colouring on maps in 1880.6 The Exhibition which Keltie and the Royal Geographical Society (R.G.S.) had arranged as a practical follow-up to his Report and which opened in London in December 1885, later travelling to Manchester, Edinburgh, Bradford and Birmingham, had revealed both the strengths and (by comparison with Germany) the weaknesses of the atlases and books available to schools. The response of the firm of George Philip was to appoint three leading geographers of the day as joint editors for ‘a new and ambitious programme’. They were John Scott Keltie, H. J. Mackinder, and E. G. Ravenstein, the latter being appointed as Geographer to the firm of Philip (1934). Many of their competitors also took up the challenge.

The need for improvement in the quality of maps and atlases was well expressed by Dr George Butler in his Introduction to the Public Schools Atlas (1884).7 ‘Hitherto,’ he wrote, ‘we have wanted good maps -- maps not too elaborate nor too costly -- to be put into the hands of schoolboys, yet containing sufficient information for the student of Physical or Historical Geography’. His atlas offered 14 maps on larger scales than many of its competitors: each double-page spread was spine-mounted so that, when opened, no folds hampered legibility. Place names were selected carefully to avoid encumbering the minds of pupils, and were precisely located with the names printed clearly and horizontally. The important physical features were rendered by hachures. The cartography was by the well-known Edward Weller. The Atlas was priced at five shillings, which seems not unreasonable for the time and for its intended market.

However, for the same price, one could buy Philip’s Atlas of Physical Geography, originally by William Hughes, which offered 40 maps illustrating ‘the natural features of the globe, the geographical distribution of natural phenomena and their connection with the industrial pursuits of mankind’. This had first appeared in 1862 and the maps were revised as new knowledge was acquired. This atlas provided clear maps and diagrams on ocean currents, relief and river systems, geological structure, climates, vegetation, animal geography, etc.. Hughes was judged highly by J. N. L. Baker8 for his view of geography as a unified subject studying relationships -- a precursor of the thesis so powerfully advocated by Mackinder in 1887. The atlas continued with maps of the distribution of mankind, the density of population, and the principal routes of maritime trade, with a descriptive notice to accompany each map.

The list of atlases displayed at the Royal Geographical Society’s Exhibition9 contained 60 titles though some of these were Bible or Scripture Atlases and others were outline maps for use in school exercises. In addition to those already named many publishers contributed items, principal among them T. Nelson and Son (Edinburgh), W. and A. K. Johnston (Edinburgh), Ruddiman Johnston (Edinburgh), Marcus Ward and Co. (London), W. and R. Chambers (Edinburgh), Blackie and Son (Glasgow), G. W. Bacon and Co. (London) and Edward Stanford (London). Too many of the atlases emphasised political geography: there was general complaint about the overcrowding of place names. One publisher, T. Ruddiman Johnston, attempted to answer this criticism by bringing out an Uncrowded Atlas of Political Geography (1886), claiming to include only the information which scholars should possess and no more and which could be easily memorised. Alas, an influential reviewer was unconvinced: ‘the author does not seem to have grasped the difference . . . . between an overcrowded map and one which, while containing nothing unnecessary shall at the same time convey to the mind of the scholar a correct notion of the general physical features of the country, the position of centres of manufacture and the chief towns’.10 Whether such criticism was an influence in Ruddiman Johnston’s decision to sell his business in 1889 to the Educational Supply Association (of which he became manager until 1895) is certainly a thought.

Reward might well be gained from following through for each of the main publishers the response made to the changing educational needs of the time. The archives of George Philip and Son, now lodged in the R.G.S., provide a potentially useful quarry. T. Nelson were certainly one firm keen to demonstrate in their Royal Map Book series (each issue at a penny or twopence, the maps and text for England bound together for sixpence) their resolve to meet the requirements of the Educational Code of 1882. Marcus Ward’s Atlas for Standard V, (1883; 13 maps), to give another example, cost 3 pence. There were atlases of the Empire ranging from John Bartholomew’s (1883) with 23 maps for 3s 6d to Keith Johnston’s Middle Class Atlas of the British Empire (no date) with 15 maps for 1s 6d. Philip’s 6d Atlas of the British Colonies and Dependencies featured Queen Victoria gazing outwards from the centre of the cover.

The 1890s brought signs of the change sought by Butler. Longmans New Five-Shilling atlas for use in Schools prompted an encouraging reaction in the Proceedings of the RGS 11 which regarded it as ‘a decided success . . . superior to any of its class yet published in this country and well suited for use in schools’. It had a distinguished editor, G. G. Chisholm. Physical features and political divisions were shown on the same map and related scales were employed. Philip’s Systematic Physical and Political Atlas (1895), intended for higher and private students, was more expensive at 10s. 6d. It was advanced by the publishers as a ‘pioneer’ work of originality and technical perfection which supplied for the first time a British school atlas more than comparable to the best atlases published abroad and giving a comprehensive survey of the new and more scientific conceptions of geography long recognised on the Continent and now being introduced into British universities.12 It was based on Philip’s Handy Volume Atlas of 1894. The work of Ravenstein and his fellow editors was bearing fruit. But John Bartholomew’s Century Atlas of 1900 (published by John Walker and Co), though meticulously produced and indexed, was more traditional, did not incorporate the ideas of the ‘new ‘geography’ and was less useful for teaching purposes.

Many elementary atlases were on the market, often in parts for the different standards of the Board Schools, but few were addressing the needs of the ‘new’ geography. Geography was now being widely and, by the standards of the time, often well taught. Classes were large, money was tight and the needs of the Codes had to be met. Bartholomew’s Elementary School Atlas (1888) was produced at one shilling to meet this need and aimed to make the subject interesting. But the rough production, lack of scales and the attempt ‘to get too much into a cheap atlas’ invited criticism.13

The breakthrough came with Philip’s London School Board Atlas, also of one shilling, of 1900. This was practically identical with the same publisher’s Elementary Atlas of Comparative Geography. The London School Board, under active scientific leadership, had supported the inclusion of four pages of maps devoted to London. There were eight pages of explanatory letter press. The experienced A. W. Andrews reviewed the atlas for the Geographical Teacher 14 and, while pointing out some deficiencies, particularly the omission of climate maps, recommended it as ‘the best school atlas yet published, apart from its remarkable cheapness’. The assistance of the London School Board’s Special Advisory Committee had been one factor: even more important, the Board, by making available copies to all its schools and by encouraging parents to buy copies at the low price, had provided a substantial guarantee. The advice of the Geographical Association had also been sought. The Atlas thus represents an early attempt in geography to bring together publishers, teachers and the educational authority to meet a perceived practical need. Suitably modified editions were later printed to provide for the needs of authorities in other parts of the country, e.g. in northeast England.

This Atlas was new (in Britain) in relating the study of maps to the area known to scholars. But there were other important innovations as the review in the Geographical Journal15 was to point out. It was ‘a great advance on the cheap school atlases published fifteen or twenty years ago’ and was appropriate to the now ‘more enlightened and improved system of teaching geography’. The scales on which the maps were drawn bore a fixed relationship to one another with the map of the British Isles (1:5,000,000) chosen as the unit. Physical features were given special prominence and a uniform system of tinting was adopted. Physical and political maps, when both were provided, faced one another. Place names were carefully selected and towns were classified by symbols according to population. ‘Hitherto,’ claimed the Preface, ‘mere indiscriminate collections of maps have done duty as "School Atlases": in the present book -- for the first time in an English Elementary Atlas -- each map takes its place as an essential part of a general scheme’. Improvements such as those introduced by the London School Board Atlas were to be found in atlases coming from other publishing houses in the first decade of the new century. For example, the London School Atlas edited by H. O. Arnold-Forster and published in 1900 by the London School Atlas Co. at 2s. to 3s 6d. (according to binding) also received an encouraging review. 16 A. J. Herbertson of the University of Oxford, the new Honorary Secretary of the Geographical Association, was associated with it and contributed ‘Introductory Notes’.

A phase of discussion on the teaching of geography and, incidentally, on the content and use of the atlas, was about to begin. Some of the better School Boards were encouraging teachers to think actively about geography and how it could best be taught. The London School Board for example held a conference on ‘The Teaching of Geography’ on November 23rd 1901.17 It encouraged teachers to run field excursions and to join teachers’ circles. It would be interesting to know more about the success of such circles: there was a brief report on the Bethnal Green History and Geography Circle in the Geographical Teacher 18 and, among others, a Hackney Geographical Circle may also have existed.

The changes that followed the 1902 Education Act and the opening of secondary schools with, in time, the appointment of geography specialists brought new opportunities,19 while the more enlightened attitude to geography in the Board of Education’s Suggestions of 1905 improved the environment for new ideas and practices.20

The Geographical Association, having widened its remit to include all types of school, had also begun to bring pressure on publishers to improve atlases and cartographical representations.21 Its journal for 1907-190822 carried the report of a sub-committee which had been established following a request from an (unidentified) County Education Authority to consider desiderata for the contents of the elementary school atlas. The sub-committee looked for an atlas of 24 pages selling at 6d, or a little more if local maps were added. Political boundaries should be shown on orographical maps. Vegetation, temperature and rainfall maps should be included. World maps were required as well as a map showing the British Empire on (note) an equal area projection. Scales should be related to one another and a uniform colour scheme employed. Only those names found in good school textbooks should be inserted. It was also recommended that the individual maps should be available for separate purchase. The next page of the same issue 23 carried a critical comment on Philips’ Model Atlas, an attempt at a six-penny relief atlas; a praiseworthy attempt but an unfortunate result!

The short but critical reviews in the journals indicate that the two societies were continuing to attach importance to the content and quality of school atlases. The Geographical Association was to take over from the Royal Geographical Society the publication of H. R. Mill’s Hints to Teaching Geography which reappeared with the help of J. F. Unstead as the Guide to Geographical Books and Appliances.24

Questions relating to the quality and use of atlases and wall maps in schools had also been raised at meetings of Section E (Geography) of the British Association for the Advancement of Science. At the Dundee meeting of the Association in 1912 it was decided to establish a ‘Committee to Inquire into the Choice and Style of Atlas, Textual and Wall Maps for School and University Use’ under the chairmanship of Professor J. L. Myres and with the Rev. Walter J. Barton of Winchester College as Secretary.25 Two subcommittees were set up, one concerned with content and arrangement, the other with style and draughtsmanship. Separate atlases, thought the committee, should be provided for senior and junior students but they should be consistent in plan and execution. A meeting was held in Oxford to consider better relationships between teachers’ needs and the practices of map publishers.

A questionnaire was devised and sent out to teachers and to Directors of Education. The committee tried to discover which atlases were in use in which types of school; for which areas was it thought desirable to have separate physical and political maps; and what ‘chief inconveniences’ had been experienced in using existing maps? An attempt was also made, using a map of Great Britain divided into areas, to establish the need for large-scale maps of particular areas. The committee also circulated for comment a preliminary draft of its report on the content of the Senior School Atlas to cover about 40 double pages, together with a list of titles.

The establishment of this committee was not as warmly welcomed as might have been expected. The Geographical Association felt affronted: it had not been asked to nominate representatives and it thought that there were too few teachers on the Committee. H. J. Mackinder was outspoken about the matter: ‘I do not deplore the competition of publishers but I do deplore the necessity of a British Association Committee’.26 There was a danger that edicts from such a body would encourage stereotyping. Trends in geography teaching had changed in a decade: a ‘definitive’ list of maps and treatments could quickly lose its acceptability.

Significantly it was John Scott Keltie who was to help to defuse the bitterness. As President of the Geographical Association he chaired a meeting at its Annual Conference at which the various sides were brought together to hear the Rev. Barton speak on ‘What should be in a School Atlas?’27

Over the past thirty years, Barton agreed, there had been ‘an immense and steady improvement’ in the subject itself and not least in the atlases designed for school use. Nevertheless, compared with what was happening in Germany, France or Italy, there was still room for disappointment with the atlases turned out by some British firms. Within the limits of an atlas to sell, say, at 3s 6d, one could not have everything. All school atlases should contain four parts. The first quarter would be devoted to world charts and maps, rainfall, temperature, vegetation, ocean currents etc. Projections, were important and, please, not Mercator. Mollweide was possible but, for himself, he preferred hemispheres. Maps of the continents would follow, with Asia, Africa and the Americas on the forty-million scale, Europe and Australia on the twenty million. Europe called for fresh thought and it was not necessary to have independent maps of France, Germany, Austria etc ‘as though political geography were the only thing in the world’. He was for regional maps, the Mediterranean, North-Western Europe, on the twenty-million scale with other maps, e.g. Italy and the Balkans, on the five million. The fourth section, the British Isles, was the most difficult. It was in the representation of their home countries that continental publishers were far ahead. In addition to the usual physical and political maps some larger scale maps were necessary, south-east England, Lancashire and Yorkshire, Glasgow and Edinburgh? And then one or two really large scale maps were essential for good teaching, the Lake District perhaps?

While agreeing with Barton on the need for maps of regions, James Fairgrieve was critical and urged the need for stronger teaching representation on the British Association Committee. J. F. Unstead was more sympathetic but had doubts on the Hemisphere versus Mollweide question. There ought, he thought, to be editions of school atlases for every locality; teaching from the known to the unknown was to be encouraged. Atlases should be reprinted every three or four years. G. F. Daniell argued for more attention to the psychology of childhood: he was also involved in the books and children’s eyesight problem. Geographers were too afraid of scale variations: children quickly understood scale. The problems as seen by publishers were explained by Mr. William Stanford: Mr. George Philip contributed as a cartographer. George Philip felt that too much emphasis had been placed on, layer colouring: this could convey wrong ideas about the configuration of the land and it should be combined with clear hill-shading. H. J. Mackinder made the point that, from the publishers’ point of view, maps were a considerable investment and a sufficiently large market had to be assured. The world maps should come at the end. Only when the students had gained some richness of thought from studying regions were they prepared for the big generalisations which are indicated on world maps’. He too recommended hill-shading, ‘Let us,’ he remarked, ‘have intelligent thought behind the cartographer’s art.’

Meanwhile, reviewers continued to point out the qualities and failings of newly published atlases. E. A. Reeves, the Map Curator of the R.G.S., approved of the new edition of J. G. Bartholomew (ed.) The Comparative Atlas Physical and Political, Meiklejohn and Son. 2s 6d, 1913 and found the physical maps to be ‘far more satisfactory than is often the case.28

But W. and A. K. Johnston’s attempt to produce a cheap school atlas, the sixpenny New Area School Atlas (1914) offered some ‘fairly good’ relief maps but others, politically coloured , are poor productions.28 Another reviewer of the same atlas29 was even more outspoken: ‘we wonder who would ever use the political maps: all geography teachers of any standing would prefer to have fewer, and would not be greatly distressed if the majority were absent. . . . [T]hey are not good examples of cartography.’ Many of the maps in Philip’s Modern School Atlas of Comparative Geography were ‘much too crudely and highly coloured and too confused for educational purposes’.30

Clearly there was still much to be done. Many schools were using obsolete atlases and lacked the funds to buy new ones. There were gaps between the expectations of the teachers, the productions of the cartographers and the possibilities as seen by the publishers. But at least the different interests were in discussion.

The Final Report of the British Association Committee appeared in 1915.31 Few changes were necessary to the lists of maps proposed in the preliminary report: it was noted that seasonal distributions of climate should be shown and that maps of the distribution of population were of high value. Other suggestions were made, as the illustration shows. However, the main part of the report provided recommendations on style, colour, lettering, paper, size of type, selection of names, projections, conventional signs and other technical matters. War-time concerns curtailed the discussion but it would be interesting to examine how far the recommendations of the committee and the responses of the teachers and publishers influenced the quality of school atlas production after 1918.


The author is indebted to Dr. Christopher Board for many helpful suggestions. The staffs of the British Library of Political and Economic Science and of the Royal Geographical Society Library have given willing assistance. Particular note must be made of the help received from Mr. Francis Herbert and Mr. David McNeill of the R.G.S. Map Room.


1. Keltie, J. S. ‘Thirty years progress in geographical education’ (1914), p. 217.

2. Wise, M. J. ‘The John Scott Keltie report of 1885’ (1986), pp. 369-382.

3. Keltie, J. S. ‘Geographical Education . . . ’ (1886), p. 451.

4. Markham, C. R. Fifty years’ work of the Royal Geographical Society (1881), p. 109.

5. Philip, G. The Story of the Last Hundred Years (1934), p. 65.

6. Freenuat, T. W. A Hundred Years of Geography (1961), p. 45.

7. Buder, Rev. George ‘Introduction’ Public Schools Atlas (1884).

8. Baker, J. N. L. The History of Geography (1963), pp. 75-76.

9. See Report of the proceedings of the Society in reference to . . . geographical education (1996).

10.Proceedings of the Royal Geographical Society, 9 (1887), p. 658.

11.Proc. R.G.S., 12 (1890), pp. 711-712.

12. G. Philip (1934), p. 66.

13. Proc. R.G.S., 10 (1898), p. 483.

14. Andrews, A. W. The Geographical Teacher, 1 (1901-02), p. 47.

15. Geographical J., 16 (1900), pp. 7-10.

16. Geographical J., 17 (1901), pp. 10-11.

17. Geographical Teacher, 1 (1901-02), p. 87.

18. Ibid. p. 118. ‘

19. Balchin, W. G. V, The Geographical Association (1993), p. 8.

20. Board of Education Suggestions for the Consideration of Teachers (1905), pp. 52-60.

21. Balchin (1993), p. 8.

23. Ibid., pp. 93-94.

24. Mill, H. R. A Guide to Geographical Books and Appliances (1910).

25. BAAS Report 1913, pp. 156-161. See also C.F. Close ‘Some Notes on Atlases and Wall maps’ (1913-14), pp. 1834.

26.Geographical Teacher, 7 (1914), p. 248.

27. Barton, W. J. ‘What should be in a school atlas?’ (1914), pp. 238-251.

28. Geographical J. (1914), p. 361.

29. Ibid. p. 362

30.Geographical Teacher, 7 (1914), p. 349.

31. BAAS Report 1915, pp. 150-158.


Andrews, A. W. The Geographical Teacher, 1 (1901-02).

Baker, J. N. L. The History of Geography (Oxford: Blackwell, 1963)

Balchin, W. G. V. The Geographical Association: The First Hundred Years 1893-1993 (Sheffield: The Geographical Association, 1993)

Barton, W. J. ‘What should be In a school atlas?’ Geographical Teacher, 7 (1914), pp. 238-251.

Board of Education Suggestions for the Consideration of Teachers and Others Concerned in the Work of Public Elementary Schools (London: HMSO, 1905).

British Association for the Advancement of Science Report 1913 (John Murray).

British Association for the Advancement of Science Report 1915 (John Murray).

Butler, Rev. George ‘Introduction’ . The Public Schools Atlas, new edn. (London: Longmans Green, 1884).

Close, C. F. ‘Some notes on Atlas and Wall maps’. Geographical Teacher, 7 (1913-14), pp. 183-4.

Freeman, T. W. A Hundred Years of Geography (London: Duckworth, 1961).

Geographical Journal, especially 16 (1900), 17 (1901), 43 (1914).

Geographical Teacher, especially 1 (1901-2), 4 (1907/8), 7 (1914).

Herbert, F. ‘The Royal Geographical Society’s membership, the map trade and geographical publishing in Britain 1830 to ca. 1930’. Imago Mundi, 35 (1983).

Keltie, J. S. ‘Geographical Education -- Report to the Council of the Royal Geographical Society’. Suppl. Papers RGS, 1 (1886)

‘Thirty years progress in geographical education’ The Geographical Teacher, 7 (1914)

Markham, C. R. Fifty Years’ Work of the Royal Geographical Society (London: John Murray, 1881).

Mill, H. R. A Guide to Geographical Books and Appliances (London: George Philip, 1910).

Philip, G. The Story of the Last Hundred Years (George Philip and Son, 1934).

Proceedings of the Royal Geographical Society, 9 (1887), 10 (1888), 12 (1890).

Royal Geographical Society, Report of the proceedings of the Society in reference to the improvement of geographical education (London: John Murray, 1886)

Wise, M. J., The John Scott Keltie report of 1885 and the teaching of geography in Gt. Britain. Geographical Journal, 162 (1986), pp. 369-382.


Paradigm Catalogue Textbook Colloquium