3 pages while Poland needed 14 pages, and the countries of the former USSR are described in merely 11 pages. The Atlantic and Indian Ocean have 10 pages each, but the Pacific Ocean is contained in 8 pages. This imbalance may have arisen by the avail- ability of data in certain areas. Thus shorter treatment may indicate some lack of information from certain areas. This could explain the un- equal treatment in its proportion to the area size of Poland. However, this will only be a deficiency to the reader who is looking for information from a specific location. Another reason for the imbalance may be her decision to shorten the text that it stays within a reasonable length. Some readers may look for temperature and precipitation records from individual locations. They will not find them in the explicit form such as in Rudloff. However, the content of station records within the climatic groups is well re- flected and interpreted in the text. The book has a comprehensive source of liter- ature by authors with climatological works. An extended subject register aids in finding individ- ual geographic places. Martyn may not consider climate classifica- tions by geographers such as Hettner, Troll, Berg, Trewartha, Bliithgen, or de Martonne as geo- graphically influenced. This may have led to her view that "there do not seem to be any works describing climates from the geographical point of view". Amazingly, the reviewer has discovered only a few printing errors in this lengthy text of 435 pages. Although the reviewer has expressed some reservations about the text it should at least be in the library of every climatologist and geographer. O. ESSENWANGER (Huntsville) Greenhouse-Gas-Induced Climatic Change: A Crit- ical Appraisal of Simulations and Observations. Edited by M.E. Schlesinger. Elsevier, 1991, 615 pp. Price Dfl.240,00. ISBN 0-444-88351-7. This volume is a collection of 35 chapters written by separate authors based upon a work- 251 shop held in 1989. The book summarizes the state of knowledge at that time on the general question of the ability of climate models to simu- late climate change induced by greenhouse gases. The book has been ably edited by M.E. Schlesinger (anyone who has contributed a chap- ter to one of his volumes will appreciate his attention to accuracy and detail as well as style) and while the chapters are written by a variety of authors with a range of points of view it comes across with a pleasingly uniform appearance and scientific level. The book presents chapters written by individ- ual authors as opposed to the consensus reports to be found in the often quoted IPCC Report (1990). Most of the main laboratories in the global climate modelling community are represented in the volume including Goddard Institute for Space Studies, United Kingdom Meteorological Office, National Center for Atmospheric Research, Geo- physical Fluid Dynamics Laboratory, the Max Planck Institute for Meteorology as well as the Department of Mathematical Modelling of An- thropogenic Impacts in Moscow. These groups report results from some of their earliest coupled ocean/atmosphere runs. In these we can begin to appreciate the subtleties in the extremely diffi- cult problem of coupling media which have such disparate time constants. Some of the most se- vero problems are not satisfactorily resolved to this day (mid-1993). Aside from the modelling progress reported there are a few chapters which look into the possibility of using past warm periods as analogs to project what a warmer world might look like. For the most part T.J. Crowley concludes that these efforts are flawed by the fact that few good analogs exist in which the forcing is exactly like that causing a greenhouse warming (i.e., the sea- sonal or other forcing structure is not similar to the greenhouse). Considerable attention is spent on examining the instrumental record: what it says, its repre- sentativeness, etc. Particularly interesting to me is a chapter by Trenberth and Olson using a climate model to simulate the errors expected in the 63 station configuration suggested by AngeU and Korshover. The section includes discussions of