2i8 DISCOVERY REPORTS 



The temperature/salinity diagram (Fig. 9) shows that E. hamata can tolerate a wide range of 

 temperature and sahnity, as would be expected from such a cosmopolitan species. Nevertheless, 

 I believe this wide degree of tolerance is more apparent than real, for in the Southern Ocean there are 

 two forms or races of E. hamata analogous to the small northern and large southern races of S. gazellae, 

 and I think that a careful study of the populations of E. hamata in the subtropics and tropics would 

 show that there also the species consists of several ecologically separate races, all of which would be 

 found to have much more precise physical and chemical requirements than would appear at present. 



In unpublished informal reports Mr J. W. S. Marr, 

 who made observations on this species for many years, 

 has shown that mixing of the two Southern Ocean forms 

 occurs seldom, but that where it does occur, as for 

 example in the Falkland Island sector of the Antarctic, 

 the mixing is indicative of similar mixing of the surface 

 waters, and that the numbers of the ' foreign ' race fall 

 off sharply on each side of the convergence, showing 

 that each race is intolerant of physical conditions on 

 the 'other' side of the convergence. The variation, dis- 

 tribution and life history of E. hamata in the Southern 

 Ocean might well form the substance for a separate 

 report on its own and may one day be thus treated. 



5'. serratodentata and S. decipiens. In the addition to 

 the typical Southern Ocean species whose distribution 

 is dealt with above there are two subtropical species, 

 S. decipiens and S. serratodentata which occur in 

 sufficient numbers south of the subtropical convergence 

 to justify the sections shown in Figs. 10 and 11. 

 S. serratodentata (Fig. 10) is often very numerous but 

 almost completely restricted to the surface layers. 

 S. decipiens (Fig. 1 1 ) is usually only present in smaller numbers and is restricted, at least in its main 

 concentrations to the 500-250 m layer. Neither species extends more than 200 or 300 miles south 

 of the Subtropical convergence. 



H. mirabilis. This species does not figure on the vertical sections and there is little to add to 

 Ritter-Zahony's description of its distribution. It is found below 2000 m., and in the Antarctic it 

 probably inhabits the cold bottom water. 



Fig. 9. The distribution of the two highest orders of 

 abundance of E. hamata in the 0° meridian in March, 

 plotted against temperature and salinity. For explana- 

 tion of the significance of the lines on the figure see 

 legend to Fig. 4. 



THE SEGREGATION OF SPECIES 



In Fig. 12 the summer distribution of all the common southern chaetognath species is indicated. 

 Inspection of Figs. 3,5, 7, 8, 10 and 1 1 shows that the rangesof mostof the species overlap extensively, 

 but if only the maximum density of each species is considered, then it will be seen that there is no 

 overlap. To show this point more clearly Fig. 12 has been constructed in which the contours of the two 

 highest densities of each species of Sagitta have been superimposed upon one another in each section. 

 In 0° and go° E five species are shown, but in 80° W only three, since two species of subtropical 

 origin are absent. It will be noticed that the main concentrations of the species do not overlap. Sometimes 

 the second highest densities overlap, but not the highest. If Eiikrohnia hamata (Fig. 8) is compared 

 with these sections, it will be seen that it does overlap the highest concentrations of some of the other 



