THELACEROS RHIZOPHORA. 559 
Tentacoli disposti in ciclo semplici—Actinine. 
Tentacoli in serie radiali quasi sempre cicoracei (fronde) di rado semplici— 
Stichodactyline. 
Tentacoli molto grandi: coperti da tentacoletti secondarii (dendriti)— 
Thalassianthide. 
But these distinctions are insufficient. It is impossible to 
regard the compound tentacles of Haeckel’s Crambactis as 
differing essentially from those of Megalactis, and in Andres’ 
own monograph ample and cogent reasons against Andres’ sub- 
divisions may be found. Again, Thelaceros possesses com- 
pound tentacles, differing not more from those of Thalassian- 
thus or of Actiniaria than from those of many Stichodactyline. 
And it at the same time possesses tentacles disposed in “ serie 
radiali.”’ The criticism of Hertwig certainly holds. It is im- 
possible to separate sufficiently the Hexactinian groups by 
tentacle form alone. The “Challenger” material did not contain 
many of these forms. Hertwig has described only Coralli- 
morphus and a doubtful Corynactis. He has pointed out the 
old distinction of Verrill—that more than one tentacle com- 
municates with the same radial chamber; but, unlike Verrill, 
he has recognised its great importance. He defines the family 
Corallimorphide as having knobbed tentacles, and more than 
one tentacle communicating with the same radial chamber. 
This group corresponds with the Corynactide of Andres if 
Hertwig’s provisional inclusion in it of Corynactis be accepted. 
But a great difficulty stands in the way. Hertwig himself 
has shown the great morphological value of the presence or 
absence of a sphincter muscle. Andres, in his account of 
Corynactis viridis, states “tentacoli retrattili; ” and Allman, 
in his original description of the genus, gives a plate showing C. 
viridis in various stages of contraction, and in one of these 
stages the oral disc is completely occluded over the tentacles. 
Hertwig distinctly states that a sphincter is absent in the form 
which he doubtfully referred to Corynactis. 
The absence of a sphincter is a prominent feature in many 
of the forms under consideration. Hertwig has shown its 
absence in Corallimorphus. It is undoubtedly absent in 
