ACTINIARIA 



very little incrusted (by spicula of sponges), provided with numerous oval nematocysts with very twisted 

 threads, 19 25 X 7 8 p in size. Mesogloea of the column thin, homogeneous, with sparse, scattered cells 

 and cell-islets. Sphincter rather strong, endodermal, with few but rough folds. Nematocysts of the tentacles 

 partly of the same kind as in the column and 17 x 7 22 X 8 (26 X 6) // in size, partly narrower and broader 

 in the basal end 22 24 X 3,5 4(5) fi, the spirocysts of the tentacles 17 X 1,5 2 to 26 X 3,5 p. Ectoderm of the 

 actinopharynx high, provided with nematocysts recalling those of the tentacles, the former 19 22 X 8 7 //, 

 the latter 23 26 X 3,5 ft. A well developed siphonoglyphe. Mesogloea of the whole actinopharynx thin. 

 Mesenteries 28, symmetrically arranged according to the macro-type, thickened in the distal part. Microcnemes 

 well developed. Longitudinal muscles relatively strong in the macrocnemes as well as in the microcnemes. 



The species, described by Verrill (1899 p. 211) as Synanthus mirabilis, is, on the other hand, no 

 Zoanthid. Though I have not seen this species, I am inclined to think that we have to do with a species of 

 the genus Stephanactis R. Hertw. (Stephanauge Verr.) = ? Amphianthus R. Hertw. = Korenia Dan., which 

 are all provided with acontia, though, according to my examination, in small numbers. The family Amphian- 

 thidae, proposed by R.He rt wig (1882), cannot be maintained. The directive plane namely is not constant in 

 relation to the longitudinal axis of the pedal disc or to the axis of the Gorgonian skeleton. Besides, I have 

 found that the mesenteries of the first order, except the directives, are fertile in these genera. I will come 

 back to these genera in the second part of this work. 



Tealidium (type T. cingulatum R. Hertw.) is a well marked genus. T. cinctum Stuck, does not belong 

 to the genus (compare below). 



According to this discussion, I think that the number of genera belonging to the family Paractiidae 1 

 must be considerably reduced. 



I have before (1893, 1898) divided the Paractiidae into two subfamilies, Paractiinae 2 and Actinostolinae. 

 The Actinostolinae is also 1893 (Nachschrift) proposed as a special family). To these subfamilies I have 

 (1918) added that of Polysiphoniinae, all based on the different development of the mesenteries. Of 

 these subfamilies Polysiphoniinae is well limited. It is more difficult to have the two former distinctly 

 separated. It is true, that it is easy to separate the typical Actinostolinae, Actinostola and Stomphia, perhaps 

 also Sicyonis from the typical Paractiinae, but as we find traces of the Actinostolid-development in such 



1 Since this was written, Stephenson (1920 p. 504) sketches the line of evolution for the old Paractiidae and the old Sagar- 

 tiidae, and derives both these families from an hypothetical ancestor, Eosagartia, at the same time dividing the Paractiidae into three, 

 the Sagartiidae into five, partly new families. I will not enter on a closer critical discussion of Stephenson's hypothesis now, but keep 

 it for the second part of this work. Meanwhile, I think that Stephenson's conclusions will have to be considerably modified. Accord- 

 ing to my statements above, the division of the old Paractiidae into three families cannot be accepted. The representatives of the 

 Marsupiferidae are Halcampids, and also the family Actinoscyphiidae must be dropped, based as it is on the presence of only 6 pairs 

 of perfect mesenteries, while the new Paractiidae should have more than 6. The genus Anthosactis namely has 6, 8 or 12 perfect pairs 

 of mesenteries. The genus, Tealidium, nearly related to Anthosactis, has at least 6 or 1 2 perfect pairs of mesenteries. Accepting Mag ui re's 

 examinations of Paranthus as correct, this genus and even one and the same species should have now 6, now 12 pairs or in the Tybec- 

 species 24 pairs of perfect mesenteries (compare above). It is evident that under such circumstances the relations between the Sagartiidae 

 and the Paractiidae will have to be seen from another point of view than that of Stepheuson. Concerning the family Sagartiidas and 

 Diadumenidae compare p. 19 and p. 21. 



1 I need not here further discuss Hertwig's formation of aspecial tribus Paractiniae for the genera Sicyonis and Polyofris, as this 

 tribus was abolished long ago, nor the affinity supposed by Hertwig between the Sicyonidae and the Tetracorallia. It is inconceivable 

 that such a well differentiated genus as Sicyonis should be nearly related to the primitive Tetracorallia. If a relationship between the 

 Tetracorallia and the Actiniaria really is a fact, it must be between such primitive Actinians as the Halcuriidae and the Tetracor- 

 allia (compare Carlgren 1918). 



