468 BOUNDARIES OF THE SEA 



have a short planktonic life, 70 per cent requiring less than 5 weeks 

 for their pelagic development. Among the species with short 

 planktonic stages are all the crab larvae examined thus far. The 

 long-distance larvae seem to fall in only three groups: the Pandalid 

 shrimps, which can also swim as adults, the Hippa or Emerita 

 group, and especially the spiny lobsters, Paniilirus. Hippas have 

 a pelagic life from 90 to 120 days, but since they often live in 

 areas where the currents move parallel to or along the coast, their 

 chances for transportation are reduced (Johnson, 1939). The 

 Paniilirus group, however, with a pelagic life of 150 to 180 days, 

 has true long-distance larvae, a fact which is in good agreement 

 with its wide and often circumtropical distribution (Chace and 

 Dumont, 1949). 



The currents, which have been considered here are only the 

 surface currents. We know that nearly all larvae are photopositive 

 just after hatching and rise to the surface. We also know, however, 

 that when approaching metamorphosis they become photoneg- 

 ative, thus approaching the bottom substratum, where they will 

 settle (Thorson, 1946; 1950). But these vertical migrations seem 

 to take place only within one and the same water mass. Most 

 larvae are, so to speak, "born into a certain water mass and seem, 

 also when transported by the current, to regulate themselves up 

 and down so as to stay more or less within it during their whole 

 pelagic life (Banse, 1955, 1956). Thus, their photonegativity in the 

 older stages will hardly stimulate them to cross a discontinuity 

 layer. If a surface layer has a thickness of, for instance, 300 m, the 

 young lar\'ae may be found at the surface, the older near 300 m 

 depth, but normally not deeper. 



For our purpose, currents which run parallel to the coastal lines 

 (the Humboldt, Benguela, Agulhas currents, etc.) are of less 

 importance. Even the Arctic currents running mostly parallel to 

 the continents and in a region where only a very few species 

 reproduce by pelagic larvae may be disregarded here. Therefore, 

 only currents crossing the oceans have been considered in these 

 calculations. 



Examples have already been given to demonstrate that a rise 

 in temperature will shorten the pelagic larval phase. This again 

 means that a current which is gradually heated on its way over 



