42 AMERICAN HYDROIDS. 



the subject, and has brought about the rejection of several genera which it would be convenient 

 and reasonable to retain. 



For instance, it occurs not infrequently in several widely different forms among the 

 Sertularida' that a gonangium will occasionally have its origin within the lumen of the hydro- 

 theca, although these species normally produce gonangia in the ordinary position. But there 

 are several species, otherwise closely related in the form of the hydrothecse and in the apparent 

 absence of the operculum, in which the gonangium normally and regularly springs from the 

 inside of hydrothecse. 



Such species should, it seems to me, be placed in the genus Synthccium, of Allrnan. 1 The 

 sporadic appearance of the gonangium of the SyntJiedivm type occurring as an abnormality in 

 other nonrelated species which commonly produce gonangia in the ordinary way does not, in my 

 opinion, invalidate the genus in the slightest degree. 



Again, we find that in many species of the Sertularidse, not otherwise closely allied, one or 

 more branches spring from the lumen of hydrothecse, although these species normally and regu- 

 larly branch in the ordinary way. But Allman has found several specimens of a certain species 

 in which the branches "invariably spring from the hydrotheca?," 2 and for this species he instituted 

 the genus Tlu'cuclddnnn. He afterwards found other colonies of the same species 3 that showed 

 the same constant character, the branches arising normally and regularly from the lumen of the 

 hydrotheca. These specimens, being more complete than the ones originally described, furnished 

 additional characters that still further confirmed his judgment in establishing the genus. Here 

 again, it seems to me, we are by no means justified in following those who would discard the 

 genus Thecocludium, because species of widely different genera will sporadically exhibit the same 

 peculiarity that is uniformly possessed by the specimens studied by Allman. 



Fourth. The number of rows of hydrothecse on the branches is a character by which groups 

 of species otherwise closely related can be segregated to form genera that appear both convenient 

 and natural. Thus we find a number of species closely related to each other and having manifest 

 affinities to certain species of Thuiaria that are characterized by having the stem beset with more 

 than two regularly disposed longitudinal rows of hydrotheca?. For such forms Allman has 

 instituted the genus Sdaginopsis.* 



Again, there are two species closely allied to the genus Scrtularclla that agree more closely 

 with each other than with other species from the fact that the hydrotheca? are placed on all sides 

 of the branch in an ascending spiral. Associated with this character in the two species thus far 

 discovered is a remarkable tendency toward anastomosis of the branches which are all in the same 

 plane, forming a fiabollate colony. For the first of these species Allman formed the genus 

 DictyoclniVunn? Another species is described in the present work. One species has been found, 

 otherwise related to the genus Sertularia, that differs from all others in the fact that the successive 

 pairs of hydrothecse are rotated on the stem so that each pair is set at right angles to the plane 

 of the pairs immediately above and below. The result is that there are four longitudinal rows of 

 hydrothecffi on the hydrocaulus, the individuals of a given row being very widely separated, and 

 each individual forms one of a pair of opposite hydrotheca 1 . For this species Allman has estab- 

 lished the genus Sftti//'r>f//ec//, 6 a genus not found in American waters. 



Fifth. The operculum can be used as an important factor in separating out some of the generic 

 groups from the great mass of Sertularidse in which the hydi - othec;e are arranged in two longi- 

 tudinal rows, but this character is not in itself sufficient, and we find it necessary to use different 

 combinations of two or more characters for this purpose, among which the combination of the 

 form of the operculum and the arrangement of the hydrothecse on the hydrocaulus is of great 

 value. There is a certain long known form that has an operculum of two valves and in which 



'Journal of the Linnu*an Society, Zoology, XII, 1S74, ]>. 365. 



2 Idem, XIX, p. 14!). 



3 ChiiUfiiiji-r Report, Hydroida, Part 2, 1888, p. 81. 



4 Journal of the Linmran Society, XII, 1874, p. 272. 



5 Challenger Report, Hydroida, Part 2, 1888, p. 76. 



6 Idem, Part 2, 1888, p. 75. 



