i8o DISCOVERY REPORTS 



but that the position in any given longitude is subject to rather more variation than the earlier data 

 suggested. 



Before the mean monthly surface isotherms in Antarctic waters can be drawn it is necessary to 

 locate the mean position of the convergence as correctly as the available data permit. Furthermore, 

 records of both the mean position and the actual position at different times and places, and the change 

 of temperature at the surface, are likely to be required for various purposes in the future, and it 

 therefore seems worth while to publish particulars of every occasion on which the convergence has 

 been crossed by the Discovery Committee's ships, and to indicate the degree of accuracy with which 

 it could on each occasion be located, and the extent of the change of temperature. 



DATA 



The 'Discovery', 'Discovery II', and 'William Scoresby' have crossed or located the convergence 

 on 139 occasions. In nearly every instance at least some indication of its position was obtained, but 

 the degree of accuracy with which it was located varied considerably according to the observations 

 taken, to the extent to which the convergence was itself recognizable, and to the angle at which it was 

 crossed. The best indication of a crossing of the convergence is a sharp change of temperature shown 

 on the thermograph (see p. 195) an instrument which was in operation at all crossings by the ' Discovery ' 

 and by the ' Discovery II ', except for a period in the fourth commission (1935-7) of the latter ship. The 

 'William Scoresby' did not carry a thermograph except on her last commission (1937-8). The other 

 important method of locating the convergence is by vertical stations taken on each side of it, whereby 

 the level of the minimum temperature of the Antarctic surface stratum can be determined. This is 

 a reliable method but is naturally of little value if the stations are very far apart. Sometimes it is 

 evident that the convergence lies between two stations at which only surface temperatures are read, 

 but there must be a clear difference, and of the correct range, if this method is to be trusted. Normally 

 the sea temperature is recorded every four hours as a routine by the ship's personnel. These records 

 are not altogether reliable as exact readings of the temperature, but in the absence of a thermograph 

 they may be of considerable value, especially when they show a sharp change of temperature at about 

 the expected position of the convergence, or when they show where the convergence lies when it is 

 known to have been crossed between two stations some distance apart. 



The best determinations of the position of the convergence are obtained when it is crossed approxi- 

 mately at right angles, and the thermograph shows an abrupt change of temperature between two 

 vertical stations of which the more southerly shows a minimum temperature above 200 m. and the 

 more northerly a minimum below 200 m. Such complete indications are not ver)' frequent, though 

 they have commonly been obtained in the Scotia Sea. In many parts of the Southern Ocean it may 

 often be found that there is no clearly defined change of surface temperature, or the change may be 

 obscured by an oblique crossing of the convergence. Sometimes two successive stations may show 

 a minimum temperature at about 200 m., and it is then difficult to know where the true convergence 

 lies. A frequent source of difiiculty is the winding course of the convergence. At the surface the 

 junction of Antarctic and sub-Antarctic water seldom lies in a straight line, probably because it is 

 an unstable boundary. It forms twists and loops that may extend as much as 100 miles north or south, 

 and it possibly even forms isolated rings. The line is perhaps comparable to the edge of the pack-ice, 

 and may be even more tortuous. The latitude in which it is found may thus depend partly on whether 

 a ship happens to cross it where it bends to the north or to the south. Another consequence is that 

 a ship steaming in a straight line obliquely across the convergence may pass from Antarctic to sub- 

 Antarctic water and back again several times. It may also happen that before or after a crossing of 



