Saturday. Go to Wakefield. I am on my way to the First International Conference on GeoComputing at Leeds University, but I am staying with a friend in Wakefield. We are supposed to meet at Kirkgate station, but I get on the wrong train at Sheffield and arrive at Westgate. The air is fresh here; either Wakefield air is very good, or that in Leicester is very poor.

Sunday. It happens to be National Heritage Weekend and places that are normally closed are open. George Gissing's house in Wakefield is open for the paying of respects. I have brought my copy of Henry Ryecroft along as a gesture, but I do not expect to do much reading.

Monday. To Halifax, to the Piece Hall. An amazing structure, which just escaped destruction in the 1960s. Is it well known? It deserves to be. There is a fine second-hand bookshop tucked into one corner. I find a copy of the National Geographic magazine for September 1984. This is the "soil" issue and contains Bob Ruhe's famous exclamation "Everybody ought to thank God for loess". I buy this for 50p and a book on mathematics for my granddaughter, which she might not appreciate.

Tuesday. Day one of the conference. It is held in the geography department, which is now housed in what I remember as the agriculture building. The conference starts well with the discovery that the proceedings are published and are in everyone's carrier bag. I find that very impressive: to have proceedings ready at the start of the conference. The first day is devoted to cellular automata, GIS on the Web, micro/macro modelling, data analysis and data quality issues. I find it all rather overpowering and sneak away to look at some bookshops at Hyde Park Corner. Thirty years ago there was a great bookshop at Hyde Park Corner; it is long gone but has been replaced by two excellent and extraordinary shops. One is accessed down some steep stairs and along a narrow tunnel, a shop for troglodytes, with a good stock of science fiction and related material. The other, across the road and round the corner, is similar: battered premises but good stock. Even some good books on the Internet, but they suddenly become out of date as I browse through them.

³ÉÈËVRÊÓƵ

Wednesday. A day for neural networks and parallel processing. I go to two good papers. One nominally concerns the making of environmental decisions with spatial process modelling, but is actually about the tussock grassland in the South Island of New Zealand. I guess you choose your title to suit your conference. I find the grassland more interesting than the methods used to study it, but it would be nice if the computer methods revealed something interesting about this tussock land which has presented a practical challenge for over a century. The slide of Erewhon station flashes past without a mention of Sam Butler. It is a serious conference.

The other paper that impresses me is from the Forestry Commission in Scotland. It concerns soil classification and related ecological problems. It illustrates a very professional attitude to the problem of growing trees in difficult country and coping with some interesting weather and challenging soils. I think I am happier with trees that grow in soil, rather than those that grow in computers. Some good posters today, and each poster session is introduced by a series of very short talks by the poster authors, which is another good idea. I like the Nottingham Trent model of soil structure. Here is one of the truly great scientific problems: how to adequately describe and model soil structure. The NTU loess model looks promising.

³ÉÈËVRÊÓƵ

Thursday. Expert systems, genetic algorithms, fractals and chaos; the geographers really have rushed into the IT age. I have heard it argued that of all the academic subjects geography is the one which has benefited most from the computerisation of life. If geography is a map, and the sheet of paper is now replaced by the computer screen, I can see the strength of this argument.

Perhaps this explains the continuing popularity of geography as an undergraduate option. British geographers do seem to move with the times. I remember in 1985 the First International Geomorphology Conference in Manchester which established geomorphology as an important independent subject. The 1996 GeoComputing conference in Leeds may be similarly successful.

Friday. Post-conference trip to Bradford, to visit the Bombay Stores, out past the university. Impressive emporium, but not really measuring up to Leicester. See two Alan Bennett playlets at Leeds Playhouse. They seemed dated and musty, concerned with the fear of computers.

Saturday. To Ilkley, in pursuit of a coincidence. As part of the Ilkley Festival an exhibition of paintings by Jack Chesterman (just retired from Leeds Metropolitan University) is being held. When we first came to live in Leeds in the 1960s we lived next to the Chesterman family in Brudenell Road. There at the gallery are all the Chestermen; we have an unexpected reunion.

³ÉÈËVRÊÓƵ

Ian Smalley

Teaches soil science in the geography department, and is an honorary research fellow at the centre of loess research and documentation, Leicester University.