GymnohJasttc Anthomedusoid and a Scorpcenoid Fish. 209 



conveyed from one certain feeding-ground to another by 

 intelligent and deeply self-interested agents. The locomotive 

 agents on their part may be sup])osed to benefit either by the 

 concealment or protection that a coat of urticating polyps 

 affords, or by the disguise, that it facilitates in the search 

 for prey. 



Well-known cases of Hydroida undoubtedly parasitic — not 

 liere to refer to Cwiina^ as only the Gymnoblastic Hydrozoa 

 are under consideration — are those of Polypodium hydrijbrme, 

 Ussow, parasitic in the eggs of the sterlet fish, and of 

 Hydrichthys mi'rus, Fevvkes, parasitic on the Carangoid fish 

 Sen'ola zonata^ Guv. 



Polypodium hydriforme (Ussow, " A new Form of Fresh- 

 water Coelenterate," Ann. & Mag. Nat. Hist. ser. 5, vol. xviii. 

 p. 110, pi. iv., translated by W. S. Dallas from Morphol. 

 Jahrb. Bd. xii.) begins its existence as a vermiform body 

 within the ovarian eggs of a species of sturgeon. Upon this 

 vermiform body primary and secondary buds appear which, 

 after five or six months of parasitic life, when the sterlet's 

 eggs escape from the ovary into the water, give rise to thirty- 

 two hydriform organisms that live free in the Volga. 



The case of Hydrichthys mirus parasitic on Seriola zonata 

 is reported by Fewkes (Proc. Boston Nat. Hist. Soc. 

 vol. xxiii. ; Bull. Mus. Comp. Zool. vol. xiii. p. 224, pis. iv. 

 and. V. ; Ann. & Mag. Nat. Hist. ser. 6, vol. i. p. 362 ; 

 'Nature,' vol. xxxvi. p. 604). The Hydrichthys colony, 

 Avhich consists of botroidal gonosomes and filiform bodies 

 (hydranths?), is attached to its fish host by a basal plate with 

 ramifying tubes. 



The filiform bodies, which are regarded as degenerate 

 hydranths, are destitute of tentacles, and the absence of ten- 

 tacles is believed to be the obverse expression of the fact that 

 the hydranths cannot catch food for themselves, and so draw 

 upon the fish as parasites. 



Mention must also be made of Gorydendrium imrasiticuMj 

 Cavolini, supposed by Cavolini, as quoted in Professor 

 Allman's monograph, to be a parasite living at the expense of 

 another Gymnoblastic Hydroid — Eudendrium racemonum. 

 But the parasitism here is doubtful. 



In the present paper 1 have to record a case of symbiosis 

 between a fixed gymnoblastic Hydroid (a species of Sty/actis) 

 and a high locomotive animal (a fish of tiic genus Miiwus)^ in 

 which it appears to mc that the association is neither accidental 



