THE ANTARCTIC CONVERGENCE i8i 



the convergence, the thermograph indicates that the ship's track passed near to a loop of the convergence 



without actually cutting it. , i • i 



It will now be seen that at many crossings of the convergence its position can only be approximately 

 determined Indeed the exact position in which it is assumed to be is sometimes little more than a 

 matter of opinion. However, even approximate positions are helpful in mapping out the mean position. 

 The latter must always be liable to modification (though in a diminishing degree) in the light of new 



'^Table 9 (p 205) is a list in chronological order of all crossings of the convergence by the Discovery 

 Committee's ships. The position of the convergence is given as nearly as it can be estimated. This may 

 be taken as the middle of a sharp or comparatively gentle temperature gradient at the surface, as hal way 

 between two stations at which vertical observations were made, etc. Since some positions could be 

 determined with a good deal more accuracy than others, an indication of the extent to which they 

 can be relied upon is given in the column headed ' Degree of accuracy'. The quality of the evidence 

 on which they are based varies considerably, and it is difficult to assign a very exact meaning to the 

 terms entered in this column. As a rule 'V. good' implies that there was a sharp rise or fall in the 

 thermograph record, clearly identified as the convergence. The possible error in the estimated position 

 is not likely to be more than about 5 miles either way. ' Good' may indicate a thermograph gradient 

 which was well defined, but which extended over some hours of steaming ; it may mean that the crossing 

 was unmistakably indicated by the ship's 4-hourly routine temperature readings; or ^^ -ay be used 

 when there was no definite change at the surface, but when two vertical ^^at-ns, not too far apart 

 showed that the convergence lay between them. The maximum error here might be ^b-^ 2° -1 s 

 either way. 'Approx.' generally indicates that the convergence may have been crossed anywhere 

 between two pofnts abo.' 50-150 miles apart; or it may be used, for example, where the only evidence 

 was a vertical station with a minimum temperature at or about 200 m., for this does not necessarily 

 mean ^h the station lay exactly on the convergence. 'Probable' generally implies that there was 

 rv^ltlhaAhe converU hid been crossed at a position which could be d-mmed^^^^^^ 

 precision but that it was not completely certain that it was in fact the convergence The erm is of en 

 ariicabk when there are several sharp changes in temperature which may indicate oops in th 

 clergenci 'Uncertain' is applied to any of the less satisfactory records, and may indicate doubt 

 as to the identification of the convergence or as to the accuracy of the estimated position. 



In normal times this work might have been undertaken by Dr G E. R. Deacon^J-J^-S- D-^^^^ 

 the war he has been in service with the Admiralty, but we have together reviewed the data on which 

 each position L Table 9 is based, and I am indebted to him for this assistance and for reading through 

 the manuscript of the paper. 



THE MEAN POSITION OF THE CONVERGENCE 

 Many of the positions in Table 9 are from isolated "bse-ations but sometinaes Ae convergence^^^^ 

 erossed more than once at positions not far apart and w.thm a few days or weeks The convergence 

 does no. appear to change its position very quickly, and such points can often be jomed up so as ,0 

 :°::a:Lr:.onofitsLrse^ta.ive„..™^^^^^^^^ 



Llralso F r, P .0.). Nos. 88 and 90 were a few weeks later and are jomed by a separate hne. 

 The tetf the convergence on these occasions may be compared wrth the mean hne ,n Plate I. 



