DISTRIBUTION 29 



123° 23' W. It will be seen from Text-fig. 10 that S. thompsoni is typical of other Southern Ocean 

 plankton in being circumpolar in its distribution (Baker, 1954). S. gerlachei, on the other hand, is 

 restricted to the Pacific, occurring with one exception in a sector between 65° S. and the ice edge 

 roughly between 175° E. and 80° W. The distribution of this species compared to other antarctic 

 species is thus remarkable in not being circumpolar. An apparent anomaly in distribution occurs in 

 the region of 150° W. where there are a number of records of S. gerlachei to the north of 65° S. Speci- 

 mens of the solitary form were taken in 1951 at stations 2743 and 2744, and as an independent check, 

 I was able to examine specimens taken on the 1949 Brategg Expedition where S. gerlachei occurred at 

 two stations which were nearly coincident in position to the 'Discovery' stations. Hydrological 

 observations in this area (Deacon, 1937, p. 31 ; Midttun and Natvig, 1957) suggest that there may be 

 a northerly outflow of surface water from the Ross Sea similar to that of the Weddell Sea current in 

 the Atlantic. Such a current system might account for the northward extension to the animals' 

 distribution that is a feature of this area. With the exception of these observations, S. gerlachei ranges 

 from the ice edge to about 65° S. and so occurs principally in an area of the West Wind Drift but 

 possibly extending in the south into the East Wind Drift. How this species has evolved and is able 

 to maintain itself in this area is not clear, since no known hydrological feature appears to serve as a 

 boundary between its distribution and that of S. thompsoni. Waters with the same characteristics of 

 temperature and salinity occur in other sectors of the Southern Ocean and yet S. gerlachei occurs only 

 in the Pacific. It is also remarkable that S. gerlachei is not taken either in the Drake Passage or in the 

 Bransfield Strait where there is interchange of water between the Pacific and the Atlantic (Deacon, 

 1937). This area particularly the Bransfield Strait is in fact frequently rich in salps, but at all the stations 

 from which specimens have been examined, only S. thompsoni occurred. 



With regard to the boundary between S. thompsoni and S. gerlachei it will be remembered (Text- 

 fig. 6, p. 20) that the change in fibre count from one species to the other is abrupt, particularly when 

 it is realized that in Text-fig. 6 the occurrences of S. thompsoni at positions south of 65° are without 

 exception from sectors other than the Pacific. Mixing of the two species is of course possible and at 

 station 1295 specimens of the solitary form of both species were taken in the same haul. They were, 

 however, easily recognized using the characters described and presumably such occurrences can be 

 attributed to localized water mixing. 



While a consideration of the quantitative distribution of S. thompsoni and S. gerlachei is outside the 

 scope of this present paper, and will be described later, it is worth noting that both species are taken 

 at certain times of the year in great numbers as, for example, at station 1261 where 66,000 S. gerlachei 

 were estimated to have been taken in a 20 min. (NiooB) haul which compares with the richest 

 hauls of S. thompsoni taken in more northerly waters. 



SUMMARY 



1. Within the 'fusiformis group' there has been a confusion of four species. Three of these have 

 hitherto been variously known as Salpa fusiformis aspera to distinguish them from the type S. fusi- 

 formis Cuvier. 



2. Using a complex of morphological characters which includes the number of fibres per muscle 

 band, the arrangement of the body muscles, and the external character of the test the four species can be 

 identified. They are A^./ww/ormwCuvier, 5. aspera Chamisso,5'.fAoOTp^om'sp.nov., and iS.^er/acAezsp.nov. 



3. Additional evidence based on variations in muscle width and stolon development supports the 

 view that S. gerlachei should be distinguished from S. thompsoni and regarded as a separate species. 



4. The southern limit to the range of distribution of both S. fusiformis and S. aspera is the sub- 

 tropical convergence. The data do not allow the northern limit to be defined. 



