6iO ME. J. E. DUEEDEN ON THE EELATIONS OP 



while tlie converse holds for forms producing a support : hence 

 we may perhaps regard the multioral maeandriform examples of 

 Micordea and Rhodactis as suggestive of an ancestry in which a 

 skeleton was present, and the last remnants of a ccenosarc in 

 Gorynactis might also be thus explained. 



Tentacles. 



The subulate form is that generally assumed by the tentacles 

 in Actiniae, but it is not unusual to find them rounded or 

 slightly swollen at the apex ; rarely, as in the Indo-Pacific genera 

 Thalassiantlius, Actiiiodendron, and a few others, they may 

 branch or become arborescent. The only species of coral I have 

 come upon with complex tentacles is Astrcea radians (Pallas). 

 In this the members of the inner cycles are all distinctly bifur- 

 cated for aboiit half their length, each half bearing a knob at its 

 free extremity. 



Though genera with subulate tentacles occur also amongst 

 the corals, the greater number possess tentacles consisting of 

 a conical stem terminated by a more or less distinct globular 

 head of a different colour. Amongst anemones, Gorynactis and 

 Corallimorphus are apparently the only forms in which the ten- 

 tacles present such a distinctly conical stem and globular head, 

 exactly recalling those of coral polyps. The stem and knob of 

 the tentacles of Gorynactis, at any rate, exhibit strongly con- 

 trasting colours. Histologically the difference is also maintained, 

 the apex constituting a battery of long narrow nematocysts, 

 while they are absent from the stem. 



The tentacles in such diverse genera as Phellia, Discosoma, 

 and Micordea, may terminate in a knobbed or rounded manner, 

 but the distinction between the stem and head is never so marked 

 as in Gorynactis and many corals. 



The arrangement of the tentacles of Rliodactis admits of a 

 close comparison with that of Gorallimorphus. In each case a 

 single cycle of marginal tentacles occurs, different sizes alternating 

 and probably representing different series not yet separated 

 centripetally ; on the disc, intermediate or accessory tentacles 

 communicate with the same mesenterial spaces as the marginals, 

 and are much divided in the first genus, but simple and knobbed 

 in the second. 



