"94 DISCOVERY REPORTS 



PART II. THE DISTRIBUTION OF SURFACE TEMPERATURE 



IN ANTARCTIC WATERS 



INTRODUCTION 

 In 1940 I published, in collaboration with Mr H. F. P. Herdman, a paper on the distribution of the 

 pack-ice, which consisted essentially of a series of charts showing the observed or estimated latitude 

 of the ice-edge in each month in all parts of the Southern Ocean. The present paper is the next step in 

 mapping the surface conditions in Antarctic waters, and its object is to estimate, as nearly as the data 

 allow, the mean positions in each month of the surface isotherms between the ice-edge and the 

 Antarctic convergence. 



During the cruises of the ' Discovery ', ' Discovery II ' and ' William Scoresby ' between 1926 and 

 1939 a very large quantity of data on surface temperatures has been collected. These data of course 

 include far more than the surface temperatures published in the Station Lists {Discovery Reports 

 vols. I, III, IV, XXI, XXII, XXIV). It is the thermograph carried in these ships which furnishes the mosi 

 important body of data, for this instrument was in continuous use (except for one short period) 

 throughout the voyages of the 'Discovery' and ' Discovery II', and in the last commission (1937-8) 

 of the William Scoresby '. v vj / y 



The data are unfortunately distributed very unevenly in time and space, for the movements of the 



ships were generally dictated by other considerations than covering the surface as equally as possible. 



u ', r ^"' ^'"""^^ ^^^'^ observations were obtained in each month, and it will be seen 



that although the Atlantic sector is well covered at nearly all times of year there are large gaps in 



other sectors, especially in the winter months. 



Some data on surface temperatures south of the Antarctic convergence are of course available from 

 other sources. These are the observations made by other Antarctic expeditions and some published 

 observations made by whaling factories. I have not, however, obtained very much assistance from 

 this material. A large number of the observations were made within the pack-ice or near its fringe 

 and It IS not certain whether some of the others are sufficiently accurate (see pp. 197-8). The number of 

 reliable records from all other sources of temperature in the open regions of the Antarctic surface water 

 are very small compared with the Discovery Committee's material, and they cover few of the gaps 

 mentioned above. ^ ^ 



Since the surface temperature in any given locality depends primarily on the time of year, I have 

 set out to plot all appropriate records on separate charts for each month, irrespective of the year in 

 which they were made. In some years of course the temperature at a given place in any one month is 

 above or below the average, but these irregular annual variations seem to have only a local and tem- 

 porary significance, and their range is much less than that of the regular seasonal rise and fall of 

 temperature. The charts used were semicircular overlapping charts identical with those used for 

 plotting the ice-edge (Mackintosh and Herdman, 1940). The isotherms were then drawn to fit the 

 actual observations as nearly as possible. Some adjustments were made after a comparison of the 

 monthly position of each isotherm, and in areas where no observations were available in certain months 

 entative isotherms were filled in by interpolation as described below. Finally, the isotherms were 

 ransferred to the circumpolar charts reproduced here in Plates II-XII. The various steps in the 

 treatment of the data are described below in some detail because they raise certain points in technique 

 which seem worth recording. ^ 



