264 THE APODID^ part ii 



fact that they fail, excepting in a few rare cases {e.g. 

 Argulus), to develop the paired eyes. The unpaired 

 eye is always present, at least in the free-swimming 

 forms ; the paired eyes appear as rudiments, only to 

 disappear again later. The characteristic caudal fork 

 of the Copepoda might well be a further development 

 of the fork which appears at the early larval stages of 

 Apus (see Fig. 41, p. 168). The characteristic ovisacs 

 may be a modification of the habit of Apus of carr}^- 

 ing its eggs about in a brood pouch, necessitated by 

 the fact that the more larval Copepoda do not develop 

 enough segments to reach the inherited place of exit 

 of the genital products, i.e. between the tenth and 

 eleventh segments. 



This theory also is quite in accord with the fact 

 that so many Copepoda are parasitic. The same 

 danger which, loosely speaking, drove the adult Apo- 

 didae into the land, and the larval to remain at the 

 larval stage, would tend to differentiate the larvae 

 themselves, as the wind has differentiated the beetles 

 in the island of Madeira.^ These insects are either 

 strong fliers or else have given up the habit of flying 

 altogether, the strong winds having swept away all 

 intermediate grades. The Copepoda are similarly 

 very markedly divided into two groups, the free and 

 powerful swimmers, and the parasites who have almost 

 or entirely given up the habit of free locomotion, except 

 in the earliest larval stages when seeking new hosts. 



The chief difficulty in the way of this derivation of the 

 Copepoda from an Apus larva is, perhaps, the form of 



^ Darwin, Origin of Species, p. 109. 



