One of the odd things about gardening in Britain, which is like nowhere else on earth in the unremitting dreariness of its climate, is that we can successfully grow a very wide range of plants outdoors, including many from regions with unrecognisably different climates. Perhaps this has more to do with the mulish refusal to accept defeat of some British gardeners than the natural propensity of, say, Juno irises to grow in the soggy conditions that we can offer them. Given our relatively impoverished native flora (those pesky ice ages) it is ironic that such diversity of plants will thrive, or at least survive, with some help. Some of the greatest challenges to extending the range of plants that can be grown well are low summer temperatures; the frequent absence of long periods of cold; stop-start springs and, at least in my case, wet soil in winter, but not hardiness. A project that I have been pondering for some time might be relevant to this discussion. It concerns the potential for using geographic information systems like Google Earth to visualise 'parallel regions'. Because GIS systems superimpose layers of different information, it should be possible (easy even) to redraw the map of the world based on whatever criteria you like. For example, if the data existed, one could ask the system to display all regions of the world lying between 500m and 1,000m asl, with a winter minimum temperature higher than -10 degrees Centigrade, with a certain minimum number of degree days in summer, overlying calcareous rocks... A fascinating extension of this idea would be to see the world geographically from the 'perspective' of a particular plant species. I suspect that this would immediately resolve many currently puzzling aspects of plant distribution. We are used to seeing distribution maps as shaded areas or dots superimposed on a map of land above current sea level. But if you're an alpine bulb restricted to basic soils above 2,000m, then the world 'looks' like a series of tiny isolated islands with huge gulfs between them. Extending the idea further, now probably into the realms of fantasy, wouldn't it be great to run the image back in time, seeing how the world view of a particular species has changed over geological time with continental drift and changing sea levels? As a way of understanding patterns of plant evolution this would be very powerful. The technology to do all of this is well established and, for all I know, is already available (please let me know if you have come across it) but I imagine that the data for some of the interesting 'layers' on my hypothetical map would be patchy at best. I'd be very interested in corresponding with anyone who thinks this sort of project might work. Best wishes, Tom