102 NATURAL SCIENCE [February 



evidently cosmopolitan,^ with characters presenting but little varia- 

 tion. The families Discosomidae, Rhodactidae, and Phymanthidae 

 ar- . • represented in the North Atlantic,- except in the Caribbean 

 area ; and while some of the species occurring there are very similar 

 to examples from the Red Sea and Australia, others exhibit im- 

 portant differences. The recent recovery in Jamaican waters of the 

 Actinoporus elegans, of Duchassaing, enables me to assert that it 

 bears no close relationship to the British Aureliania, as was sup- 

 posed by Andres. Its anatomical structure shows it will probably 

 necessitate the erection of a new family. 



One important distinction between the Actiniae of the eastern 

 and the western hemispheres can now be made in regard to the 

 Thalassianthidae, Cryptodendridae, and Phymanthidae, families with 

 arborescent or warted tentacles. So far as known, American 

 waters are devoid of genera such as Heterodactyla, Thalassianthus, 

 Megalactis, and Actinodendron, all bearing more or less com- 

 plicated branching tentacles, while they are characteristic Indo- 

 l*acific types.^ The only West Indian species with well developed 

 arborescent organs is Lebrunea neglecta, and these are columnar not 

 tentacular in their origin ; the frondose areas of the Phyllactidae, at 

 one time regarded as tubercular and papillose tentacles, have been 

 rightly compared by McMurrich with the acrorhagi of the Bunodidae, 

 and are therefore also columnar. 



Comparing the polyps of one area with those of the other the 

 gradation from the simple to the complex tentacle can be traced in 

 several genera and families. It is well illustrated by the genus 

 Phyriianthus. The West Indian P. criicifcr has mere simple or bilobed 

 tubercles along the oro-lateral aspect of its tentacles ; indeed, as I 

 have frequently found among the abundance of specimens at the 

 Port Royal Cays, even these may be entirely absent, or individuals 

 may be obtained representing all conditions in their numerical de- 

 velopment. The tentacles of P. loligo, of the Red Sea, on the other 

 hand, present us with all stages from a mere tubercular to a dendritic 

 condition ; while the so-called disc tentacles may be also branched, a 

 condition which never occurs in the Western species. What may 

 be regarded as the final stage of the series is exhibited by P. 

 mucosus from Torres Straits, where Haddon found the lateral ap- 

 pendages to be even more dendritic than in P. loligo. The mar- 

 ginal and disc tentacles of the Caribbean Rhodadis Sancti-Thomae 

 should be similarly compared with the irregularly palmate or 



^ Draytonia mijrciu, D. and M., now obtained in Jamaica, has proved itself to be a 

 Corynactis. 



^ Dr O. Carlgi-en, in his ^' Nordische Actinicn," mentions no Stichodactylinae nor 

 any Zoanthidae. 



^ Kwietniewski has just described a Thalassianthus and a new genus of Discosomida* 

 from Ternate, one of the Moluccas Isles. 



