^^Q DISCOVERY REPORTS 



swimming power. Whatever the cause, there is little doubt that the depth attained during migrations 



increases with age. 



To the south of the Antarctic convergence the larval stages are thus confined mainly to the north- 

 easterly moving Antarctic surface water and the effect of this can be seen in their concentration in the 

 region of the convergence. 



SWARMING 



There are no records of E. triacantha forming shoals on the same scale as E. superba (Mackintosh and 

 Wheeler, 1929; Hardy and Gunther, 1935 ; Marr, 1956) or such as are sometimes seen in Meganycti- 

 phanes norvegica (M. Sars), 1857 (Macdonald, 1927). On the contrary, the adults are generally rather 

 sparsely but evenly distributed. Out of 292 night hauls between 100 m. and the surface with the 

 N 100 B only ten contained more than fifty adolescents and adults. In nine of these the majority were 

 juveniles in the post larval stage and in the tenth over 90% were under 20 mm. Thus, although 

 swarming does not occur in the adults, occasionally hauls have been taken which contain numbers of 

 the younger stages that are significantly larger than average. Two of these are of particular interest— 

 at St. 1936E, one of a series of stations repeated at approximately 3-hourly intervals, and at 

 St. 2807. At St. 1 936 E a flight of five i m. nets were towed horizontally at the following depths: 

 5, 22, 44, 86 and 133 m. The upper three nets contained o, 517 and 8 larvae, respectively. These 

 larvae must have been confined to a layer not more than 40 m. in thickness and very probably much 

 less since the smaller number of larvae in the 44 m. net does not necessarily indicate a continuation of 

 the concentration of larvae down to this level, since this net fished for a short time at 22 m. while the 

 net above was being attached to the warp. It is not possible to determine the horizontal spread of this 

 group of larvae owing to their diurnal vertical movements, but at St. 1936D worked half a mile to 

 the south two hours previously a similar flight of five nets fished at intervals down to 155 m. took 

 a total of only forty-eight larvae. 



At St. 2807 an N lOoB was fished from only 33 m. to the surface owing to bad weather, but even 

 in this shallow haul 248 larvae were taken— a relatively dense swarm, at least for E. triacantha. 



Marr (1956) finds that the older stages of E. superba accumulate in thin plate-like shoals and, 

 although there is no evidence to show the horizontal extent of the patches of E. triacantha larvae that 

 have been found, it seems probable that they too only occupy a comparatively narrow vertical range. 



DEVELOPMENT AND GROWTH 



Larval stages 

 The most comprehensive account of the larval stages of Euphausia triacantha is that by John (1936, 

 p. 278). In this he describes and figures the second and third calyptopes and the seven furcilias. The 

 first calyptopis is described by Rustad (1934, p. 19). Up to the present time no larval stage earlier 

 than the first calyptopis has been found which can be referred with certainty to E. triacantha. Ruud 

 (1932, p. 55) describes a metanauplius with a denticulate margin to the carapace which he suggests 

 probably belongs to either E. frigida or E. triacajitha ; many metanauplii similar to this have been 

 found in the ' Discovery ' material, but as these have been taken not only within the range of E. tri- 

 acantha but also to the north and south it seems certain that this type of metanauplius occurs in more 

 than one of the southern species of Euphausia. 



Table 7 shows the range and mean lengths of the ten larval stages examined ; these data are plotted 

 in Fig. 13. Measurements are from the anterior margin of the eyes to the tip of the telson. The mean 

 length of the seventh furcilia is probably rather low, for there is tendency for the second pair of posterior- 



