THE number of persons requiring information about the temperature normally experienced in different parts of the British Isles, for one purpose or another, has for many years been large enough to make it an important part of the work of the Meteorological Office to secure so far as possible that standard methods of obtaining air temperature shall be followed both at official and private meteorological stations, and that summaries of these records in comparable form shall be available for inquirers. In a recent handbook (“Averages of Temperatures for the British Isles.” H.M. Stationery Office. 9d., postage extra) monthly and annual averages of the daily maximum and minimum temperature are given, so far as possible, for the years 1901-1930. As there are, however, many stations for which the averages can refer to only a portion of that period, the inquirer has to be warned against indiscriminate comparisons; for example, differences between a pair of stations for which the period of years referred to is not the same may be due more to peculiarities of the two periods than to real climatic differences; there is the further pitfall of possible differences in times of setting of the maximum and minimum thermometers at the two places. These matters are dealt with in the introduction and the necessary information is shown against each set of figures. It may be observed that the precise meaning of the ‘normal’ or ‘average’ maximum or minimum temperature for a given season and a given place is not easily defined. In Table I of the “Book of Normals”, which the tables under review supersede, the mean temperature at Kew in January (or rather the mid-point between the mean daily maximum and mean daily minimum) is given as 38-9°F., whereas in the new tables it appears as 40-4° F. The relatively low figure in the “Book of Normals” is due partly to the fact that in the period covered therein (1881-1915) there was a notable run of cold winters, those of the early ‘nineties, while the winters of 1901-30 have mostly been mild; there is no means of knowing whether the next 30 years will give an average or ‘normal’ near to 38-9° F. or one nearer to 40-4°.