2IO DISCOVERY REPORTS 



never been determined in that part of the eastern Pacific. The presence of a few specimens of 

 S. serratodentata at the northernmost station of the line (1320) suggests that the subtropical conver- 

 gence might be within 200 miles or so of that position, for the species is found elsewhere in the 

 subantarctic only within that distance of the subtropical convergence. The isotherms off the coast of 

 southern Chile run parallel to the coast (Gunther, 1936) and this suggests that there may be a south- 

 ward loop of the convergence lying from about 45*^ S 90'' W to about 53° S 80° W and thence 

 northward in the meridian of about 77° W (see Fig. i), but there is no hydrological information other 

 than that mentioned to back up this view. In 80° W the subantarctic zone extends to a fairly sharp 

 convergence in 63° 30' S and the antarctic zone was accessible thence to 68° 18' S. 



In view of the differing length of the three lines, the antarctic convergence has been used as a point 

 of reference in the diagrams. 



For making the sections shown in Figs 3, 5,7,8, 10, 11, 13 the data given in Tables 4 a-c (p. 225) were 

 used. The contours adopted indicate the following quantities in terms of numbers per 250 m. haul : 



It will be seen that the contours are not the same for all the species, and this is due to the very large 

 differences in total numbers of the species present. Originally the contours were chosen by a simple 

 progression, e.g. 5-10-20-40, but it was found that in certain species strict adherence to this method 

 produced rather confused-looking sections, and a certain amount of allowance had to be made. In 

 S. maxima and S. marri, for example, it was not felt necessary to indicate occasional stray specimens 

 obviously separated from the main population, and as a result the contours for these two species begin 

 at 4 and 5 respectively. In S. maxima, examination of the data showed that contours at 20 and at 40 

 would be an additional complication which would show nothing more than a single contour at 30. 

 It should perhaps be pointed out that more contours tend to emphasize the separateness of the species 

 shown in the maximum density sections in Fig. 12, but the extra amount of lines and shading tend to 

 make the figures look confused. S. serratodentata has a relatively compact population and when it is 

 present in large numbers they are really large, e.g. 620 and 715 for single hauls in 0° and 90° E, and 

 it was not thought necessary to indicate any more than the contours given. 5. decipiens also shows 

 a similar compactness with smaller numbers. E. hamata is often present in large numbers, and it was 

 not thought that contours between 20 and 100 would show anything better than the two at those 

 densities. 



For the winter sections where the numbers of all species were lower, lower contours were used and 

 these have followed the simple progression except in the case of E. hamata, where again the inter- 

 mediate contours seemed to be unnecessary. 



S. gazellae (Fig. 3). The northern limit of the species is shown to be just north of the subtropical 

 convergence in 0° at a depth of 200 m. and probably it is at a similar point in 90° E though at less 

 depth, but in this line there is no negative station north of 2802 to prove it, though there is little doubt 

 that the northern limit is near the convergence (David, 1955). In 80° W it is evident that the species 

 extends well north beyond Station 1320 which is well south of the subtropical convergence. The 

 southern limit in 0° is probably just at the continental slope, though there is not a negative station at 

 the southern end of this line. In 90° E the species does not extend nearly so far south but ceases in 



