THE SIZE OF SOUTH POLAR ASCIDIANS 153 



the three species now seems to be complete, although they occupy adjacent areas. Ecological dif- 

 ferences also exist and the percentage distribution of the ' Discovery ' specimens at different depths is 

 given in Table 39, showing that the more southern species inhabit progressively greater depths. 



Table 39 



Depth {metres) 



More than 

 0-100 100-200 200 



P. legumen 100 o o "j 



P. georgiana 18 64 18 \ Percentage distribution 



P. bouvetensis o 12 88 J 



The same family provides another example of geographical speciation. Pyura stubenranchi and 

 P. setosa appear to represent respectively the Subantarctic and Antarctic species derived from a 

 common stock, and their geographical ranges do not overlap. 



In the family Styelidae there are also instances of the same process. The little-known species Styela 

 schmitti shows slight but apparently constant structural differences in specimens from Montevideo 

 and from Cape Horn, which I believe to represent geographical forms. As the species has been 

 collected only twice, however, it is not possible to say with certainty whether its distribution is con- 

 tinuous, with the observed anatomical differences representing different points of a cline, or whether 

 it is discontinuous with separate geographical subspecies. 



I have already mentioned the difficulty of deciding if Cnemidocarpa nordenskjoldi and C. drygalskii 

 are different species. If they are, and I have given reasons for thinking them to be separate, they 

 constitute a pair of very closely related species, one species occurring only in the Subantarctic and 

 the other only in the Antarctic. 



The family Molgulidae, as represented in the ' Discovery ' collections, gives one example of popu- 

 lations of a species from different areas showing structural differences. This is Molgula setigera, with 

 distinct forms from South Georgia and Marion Island and the typical form from the Falkland Islands. 



Conditions in the Antarctic and Subantarctic seem to have favoured the geographical speciation of 

 ascidians. The same has been found in other groups (see Ekman, 1953), and has been explained by 

 the long geological isolation of the Antarctic Continent. The position of the Antarctic Convergence, 

 coinciding as it does with the barrier formed by deep water, must have been effective in maintaining 

 the separation of populations upon which geographical speciation depends. 



The presence of isolated islands near the Antarctic Convergence also encourages the breaking up of 

 populations into groups which become genetically isolated as subspecies or species. 



THE SIZE OF SOUTH POLAR ASCIDIANS 



The size attained by an ascidian depends mainly on the water temperature and the food supply. In 

 a species adapted to cold water it is the food supply which limits growth. Ascidians feed largely on 

 plankton, which is abundant in Antarctic waters, so it is probable that the rich phytoplankton is the 

 primary cause of the large size which certain south polar species attain. 



A survey of the ' Discovery ' material gives little evidence of an increase in the size of specimens of 

 one species from areas nearer the south pole. In fact not many species cross the Antarctic Convergence 

 so it is difficult to find species in which this tendency might appear. Aplidium circumvolutum, however, 

 does increase in size towards the pole. But in Distaplia cylindrica, which occurs in both the Sub- 

 antarctic and the Antarctic, almost equally large specimens were taken from South Georgia and the 

 Patagonian Shelf. 



