A 



f 



130 



THE OPIIIURIDiE. 



evidently not all closely related. Thus, 0. aurantiaca Verr. and 0. inermis 

 Lym. are not very, nearly related to 0. deshayesi Lyrn. and O. abyssorum 



' ^ 'rf 



Lym. y again 0. maculala Ludw. belongs to the group of 0. ciliata and has no 

 close relation to any of those named above. 0, fraterna Lym. and 0, sau~ 

 riira Verr. have no arm-comb either ; but as these species are not figured, 

 it is almost impossible to say anything with certainty about their relations. 

 A character, which may be of importance, is to be derived from the 

 tentacle pores : whether they continue far out on the arms, or are only 

 found on the first four-five joints, as in our 0. nana and 0, divisa^ and without 

 doubt in several of the species figured by Lyman". The most important char- 

 acter however seems to be the form of the ventral plates. Li Lyman's first 

 group (type O. ciliata Retz.) the ventral plates are short, all widely separated, 

 much wider than long beyond the second plate ; in his second and third group 

 the ventral plates of the inner part of the arm are as broad as long, almost 

 hexagonal, in contact in their whole breadth. Tliese two groups are dis- 

 tinguished from each other by the length of the arms : in the second 

 group (type 0. nodosa Ltk.) only twice the diameter of the disk; in the third 



r 



(type 0. bullata Wyv, Thorns.) much longer, about 4-5 times the diameter 

 of the disk. To the latter group belongs our 0. supcrha. Some of the 

 species, however, which Lyman refers to this group, do not appear really 

 to belong to it, e.g. 0. ornata Lym. and 0. radiata Lym. These would per- 

 haps better form another group with 0. radiata Lym. as type, characterized 

 by the axe-shaped ventral plates. To this group would belong our O, ahcisa 

 and perhaps 0. ohtecia. Another group apparently would be formed of the 

 species with ventral plates, as in our 0. tiimidosa and 0. plana. Finally, the 

 specimens wath all the ventral plates widely separated, which do not belong 

 to the first group, would perhaps form one or two special groups. Evidently 

 a natural subdivision of the genus OpfdoglypJia would not be very difficult to 

 work out, but it will be necessary to have ample material for comparison.. 



Ophiocjljjpha Forhesi Duncan is said by Koehler (128) to form an inter- 

 mediate type between the genera Opldoglypha and OpfMoniiismm, That 

 surely is not correct. The most important character which distinguishes 

 Ophiomusium from Oj)hiogl//pha is found in the first pair of tentacle pores 

 (in reality the second pair). In OpJuogl//pha these pores are very large and 

 conspicuous, in Op)Momiismm they are not at all seen from without, but lie 

 totally in the mouth space. (This character however was not mentioned by 

 Lymak.) Only Ophiomiismmjlahellum Lym. ^nd . pulchcUum W yv . Thorns. 



i 



■'^i 



M 



» 



f. 



\ 



