DISTRIBUTION OF THE CHAETOGNATHA 203 



brought to the surface, and which do not catch anything while being lowered, far outweigh the slightly 

 erratic manner in which the larger animals are caught. In so far as the Chaetognatha are concerned 

 it is only the large specimens of Sagitta gazellae and S. maxima which are to some extent missed by 

 the nets, the other species being small enough to be adequately sampled. 



The results, corrected to represent numbers per 250 m. haul (Mackintosh, 1937, p. 377), have been 

 plotted by contours of density of numbers in the manner of a physical section for temperature, 

 salinity, etc. This method of plotting has many limitations; the stations may be widely spaced and one 

 may question the legitimacy of presuming that any population of animals is continuous over the 

 hundred miles or so between stations. When there are small numbers of animals, this objection is 

 probably valid ; yet the consistency of the pattern of distribution shown by the sections (Figs. 3-8) is 

 surely evidence that the method of presenting the data is not badly at fault ; for if the populations 

 consisted of randomly spread, small patches of animals, it is unlikely that such patches would be found 

 at two or three stations in such similar numbers that they could be plotted by the contour method with 

 the results shown, and further, even if the populations were patchy but not randomly spread, as is quite 

 likely, then the contour section will show the area of abundance of such patches, which is, after all, 

 what is required. One disadvantage of the use of nets, such as the N 70 V, which fish through 

 relatively wide horizons, is brought into prominence by the use of the contour-plotting method; that 

 is, that a haul containing a large number of two or more species has to be plotted as if both species 

 occupied the same place in the water column, yet it is quite likely that on occasions when the net has 

 passed through two (or more) water masses in one haul the plankton populations are as separate as 

 the water masses themselves. 



No subsampling has been done except possibly in the 80° W line where the numbers given by 

 Mackintosh (1937) have been used. In this line, however, the northernmost stations have been re- 

 examined for subtropical species which might have been missed in the original sorting which was done 

 in conditions unfavourable for the careful microscopical examination required to identify, for example, 

 very young specimens of Sagitta serratodentata. The fact that some of these have now been found does 

 not in any way affect the conclusions reached in that paper. 



In addition to these lines of vertical nets, consideration of other vertical and oblique net data has 

 been necessary to describe the distribution of the rare and the exotic species which are sometimes 

 found in the Southern Ocean but which did not happen to occur in the four lines of stations used to 

 describe the distribution of the commoner species. 



A SUMMARY OF THE SPECIES OF CHAETOGNATHA FOUND 



IN THE SOUTHERN OCEAN 



The chaetognath fauna can be conveniently considered as composed of three elements : (i) ' endemic ' 

 species, (ii) species common to other regions which maintain themselves in the Southern Ocean, and 

 (iii) exotic species which probably do not maintain themselves in the area but which are carried in 

 presumably by water movements from other regions. In addition there are a few species reported by 

 various authors as occurring in the area, but which have not been found in the extensive Discovery 

 collections and which are probably the result of misidentification or mistakes in labelling. 



I. 'Endemic' species 

 {a) Sagitta gazellae 

 Sagitta gazellae Ritter-Zahony, 1909. Die Chdtognathen der Gazelle-expedition. Zool. Anz. xxxiv, pp. 787-93. 

 S. gazellae Ritter-Zahony, 1911: Germain, 1913; Jameson, 1914; Burfield, 1930; Bollmann, 1934; David, 1955. 

 S. hexaptera Steinhaus, 1900; Fowler, 1907; Germain, 1913. 



