CERTAIN STICHODACTTLIN^ TO THE MADBEPOEARIA. 641 



McMurricln (1889, p. 46) states that the marginal tentacles 

 in JRicordea are in a single row, but Jamaican examples are 

 dicyclic, the outer row corresponding with the disc tentacles and 

 the inner and larger alternating. The coloration of the stem 

 and rounded head is different. 



With regard to the close external resemblance borne by 

 the tentacles and other parts o£ Gorynactis to those of a coral 

 such as Caryophyllia, the opinion of Grosse (1860, p. 292) may be 

 here quoted : — " There is much in the appearance of this animal 

 [C. viridis] which agrees with CaryophylUa : the colours and 

 their distribution, the general translucency of the tissues, the 

 form and crenation of the mouth, and, in particular, the shape, 

 arrangement, and minute structure of the tentacles, are so 

 exactly those of the Coral, that I have often more than half 

 suspected that the former is the immature condition of the 

 latter." 



According to Gosse (p. 309), the tentacles in CaryophylUa are 

 " set in several rows, diminishing in size from the outer row 

 inward, each consisting of a stem with a globular head " ; an 

 arrangement corresponding with that in the stichodactylinous 

 Gorynactis. In the development of Garyophyllia cyatlms, Gr. von 

 Koch (1897, p. 759) found the tentacles in the first two cycles 

 of six to alternate. 



With the exception of odd irregularities the tentacles in 

 Gorynactis viridis and G. myrcia are tetramerous. Such an 

 arrangement may occur at times in other anemones, but is 

 apparently never so constant a feature as in this genus. My 

 observations show that the mesenteries in G. myrcia are likewise 

 in fours, eight complete pairs representing the first and second 

 cycles, and eight incomplete alternating pairs representing the 

 third. This tetrameral symmetry in an anemone possessing so 

 many coral characteristics suggests the consideration of its 

 relationship with the ancient " Eugosa." Connected with the 

 question should be borne in mind the interesting fact that 

 in the octamerous coral Garyopliyllia rugosa, Moseley, Gr. von 

 Koch has shown (1889) that at first six septa of the first order 

 are developed, then six of the second order, and that it is only 

 with the appearance of the third order that the development 

 becomes altered in such a way that we ultimately obtain the 

 octamerous condition of the adult. 



