226 DISCOVERY REPORTS 



In general it will be seen a single developmental phase (or at most two adjacent developmental 

 phases) makes up the vast bulk of any larval sample, many of which indeed, the majority in fact 

 of those from the deeper levels, consist of larvae all, or virtually all, at the same fixed stage of 

 grovrth. It will be seen, too, that the stage frequency throughout is normal, the mode, in so 

 far as a single stage may be said to represent a mode, in almost every instance being sharply 

 defined and of high, in many instances of the highest possible, value. The number of stages 

 found together in any one sample rarely exceeds four and when five or more are encountered 

 this happens only at levels above 250 m. In the deeper strata (Figs. 32 and 33) it is rare in fact 

 to find more than two, the samples from below 500 m. being represented exclusively by one, or 

 by one and by one other immediately succeeding larval stage. The position, however, in the 

 500-250 m. layer is perhaps the most striking, for there it will be seen in all but three instances 

 the samples consist exclusively of First Calyptopes, a phenomenon that springs from the fact that 

 the metamorphosis of the Metanauplius (p. 97) is accomplished well below the 500 m. level, 

 the resultant First Calyptopis during the final phase of the developmental ascent traversing the 

 500-250 m. layer without moulting except in such rare instances as for example are provided 

 (p. 90, Table 13) by the enormous concentration of larvae encountered at Station 1138. The samples 

 from the upper strata (Figs. 34-7) are also represented quite frequently by one stage alone 

 although there just as frequently by two, three or four. The frequency of occurrence of multiple stage 

 samples, however, would naturally be higher in the upper than in the lower strata, for it is at the 

 higher levels near the surface that the larvae first begin actively to feed, and to grow and moult on 

 a massive scale. 



The dominance, repeatedly encountered in our samples, of one, or of one and one other adjacent 

 larval stage, and the high modal values that go with it, can only it seems have one explanation, namely, 

 that the patchiness of the larvae that persists from their earliest deep appearance until the completion 

 of their life-cycle in the surface, arises directly from their being assembled in a system of swarms in 

 which the separate individuals in each instance are of the same or much of the same age and which 

 from hatching onwards seem to keep together until the end of their natural span. There is nothing 

 new in such an idea. Gurney (1924), for instance, in his account of the Decapod larvae of the 

 British Antarctic ('Terra Nova') Expedition, writes, 'It seems probable that the larvae of Decapods 

 and Euphausiids are not so much at the mercy of the currents as might be supposed. It is not very 

 unusual to find swarms of the larvae of one species in different stages of development, which seems 

 to indicate a power of keeping together from hatching onwards, or of collecting in a suitable locality '. 

 Again, Russell (1927a), recording a dense swarm of Corystes cassivelaunus zoeas in the Plymouth area, 

 and 6 weeks later and in the same position a ' great shoal ' of megalopas of the same species, writes of 

 the latter, ' It would be interesting to know whether this was the same shoal that was met with on April 

 8th, when the Corystes were then in the later zoea stage. It is, at any rate, an indication that the larvae 

 must have some considerable powers of keeping together if, when they have passed through all 

 the stages of their pelagic existence, they can form so large a shoal as that met with on May 19th, 

 1925 '. What these powers could be, unless it simply be that these creatures, having originally been 

 brought together through the hatching of dense concentrations of eggs, prefer thereafter always to 

 keep in close contact with their own kind, will perhaps always be a matter for conjecture. As 

 Bainbridge (1953) remarks, 'there may be many means by which animals keep together in swarms, 

 such as sight or a rheotactic sense', while Hardy it will be recalled (p. 154) suggests there 

 might be some ' guiding principle — almost as if there were some leader in command of the whole ' 

 which keeps the swarms intact, and it may be that this principle is pure instinct, 'an inborn 

 capacity for doing apparently clever things, suffused with dim awareness and backed by strong 



