THE OSTEACODA. 223 



specimens. The further that we go back in the ontogeny, the more do the organs in 

 question resemble one another ; even the shell-contours of different species tend to 

 converge towards a common and, presumably, ancestral type. Dr. Miiller made no 

 attempt to define the characters of the early stages, although they can hardly fail to 

 have been present in large numbers in the collection. 



In handling large collections from tropical or subtropical waters (such as those in 

 which the ' Valdivia ' very largely worked), owing to the great number of species present 

 in a havil and to the scattered nature of the stations, it is far more difiicult to make due 

 allowance for such perturbing factors as variation, developmental changes, and sexual 

 dimorphism, than is the case in temperate or sub-polar waters ; for in these latter the 

 large number of specimens but a comparative paucity of species are more characteristic. 

 In other words, the argument, that two very similar forms (whether male and female, or 

 larva and adult) are likely to belong to the same species because they are constantly 

 associated together in the collections, is far stronger in temperate than in tropical seas. 

 Probably all who have worked over large plankton collections from temperate waters 

 Avill agree, that very closely allied species — species studied in the detail which has 

 become necessary in such groups as Copepods, where a few hairs more or less may turn 

 the scale — are rarely found together at the same horizon (sea-climate) *. In tropic 

 waters this is apparently not so marked. Too much stress must not be laid on what may 

 be called " the argument from constant association," but it is no doubt of considerable 

 value in dealing with southern or northern collections su.ch as the present. 



Again, the piu'ely local nature of the present collection is a great help ; in the com- 

 paratively scattered collections of the ' Valdivia,' which probably met a fresh combination 

 of species into the fauna at nearly every station, it would have been almost impossible 

 to have worked out the relations of larva? and adults ; in the small area studied by the 

 ' Research,' where the general combination of species to form the local fauna hardly 

 varied from day to day, it is far easier to connect together successive stages of the 

 same species. 



Consequently an attempt has been made in this Report to link up larvae and adults 

 wherever practicable. Until this has been done, it is dangerous to use Ostracoda as 

 covinters in a discussion of problems of oceanic distribution. For, although the younger 

 stages of many organisms appear to react to their environment in a manner dilferent 

 from their adults, still this cannot be taken for granted in every case without proof. 



As in the case of some other groups already described, I have tried to sort out from 

 the general collection every specimen of an Ostracod captured. With the minute 

 specimens of less than 1 mm. in total length, this is of course rctiUy impracticable. 

 But, as I have already pointed out (pp. 7, 93), the error will merely underestimate 

 the difference between abundance and scarcity, and is therefore an error on the 

 safer side. 



The difl&culty in connecting immature larvae with their sexual adults in Ostracods and 

 other Crustacea, lies in the alteration of the form of the shell and appendages at each 



* An apparent instance, however, will be found below in Conchcecia magna and zetetios (p. 310). 



