HYDROIDA II 63 



however, be admitted as a particularly important feature in generic distinction, as will be seen from a 

 closer study of the Nemertesia species and their development. A species such as Nemertesia tetrasticha 

 (Meneghini) differs in reality but little from the pure Phimularia type, its four-rowed appearance being 

 due to a slight alternating displacement of the apophyses on the two sides; we have here then, as a 

 matter of fact, two-rowed colonies, which have probably when younger been of the pure Phimularia 

 form. In Nemertesia ramosa (Lamarck) I have in a previous work (1912 p. 5) pointed out that the distal 

 parts of the colony not infrequently exhibit a purely biserial arrangement of the hydrocladia; the 

 present material shows that small colonies both of this species and of Nemertesia antennina (kiune) 

 are altogether biserial. Here, as a matter of fact, it is quite another character which warrants the 

 generic identity, to wit, the large sessile sarcotheca ("mamelon") of the apophyse, which occurs in the 

 colonies from the earliest stages onwards. 



The development of the colonies, and the grown colonies themselves, clearly show that the 

 multiserial arrangement of the hydrocladia is a secondary adaptation which is, moreover, in many 

 species, subject to great variation within one and the same colony. The number of branch series is 

 even variable in the single species, as in Nemertesia ramosa, where it varies from two to eight, or 

 Nemertesia antennina, where we find from four to twelve. It is not easy to see why it should be 

 more reasonable and effective to draw the limit between biserial and multiserial than for instance be- 

 tween quadriserial and multiserial. On the other hand, there is one constant feature in all the species 

 hitherto noted under this genus; to wit, a peculiar large sessile sarcotheca on the apophyse. A com- 

 parison with other species (see for instance more particularly Aglaophenopsis cornuta) suggests that 

 this formation should in reality be regarded as an abortive hydrotheca. In most cases, it forms a 

 rounded raised portion, with circular opening, happily described by the French writers as a "mamelon". 

 This sarcotheca occurs in colonies from the smallest to the largest in the same manner, whether the 

 species is biserial or multiserial. It must thus be regarded as of generic importance. Consequently, 

 we must reject from Phimularia those species which are furnished with a "mamelon" on the apophyse 

 and have unbranched hydrocladia, placing them instead under Nemertesia. 



We thus obtain, in this genus, biserial and multiserial forms. It is impossible to determine, 

 from the data available, whether a species like Nemertesia caulitheca (Fewkes) (= Phimularia cauli- 

 tl/eea, Nutting 1900 p. 63) is a young stage or not; its gonaugia have not yet been discovered, and it 

 may possibly be found to go through a development similar to that of our northern species, passing 

 0-radually from a biserial type over into the multiserial form. If not, then it must be regarded as a 

 primitive species. The next stage is represented by Nemertesia tetrasticha (Meneghini), with its alter- 

 nately displaced apophyses in the two branch rows; Nemertesia ramosa undergoes this stage, and 

 probably also Nemertesia antennina, the former, however, at any rate soon proceeding further with a 

 marked twist of the internodia, so that the apophyses in one internodium are perpendicular to those of 

 the previous and subsequent ones. Not until later do further apophyses appear at the same level as 

 the first, and from now onwards we regularly find, owing to the revolved position, a number of branch 

 series divisible by two. 



The nature of the stem has also served as basis for a division of the genus; here again, how- 

 ever, there is no real difference of principle to work on. The younger stages of Nemertesia ramosa 



