4 PROCEEDINGS OF THE NATIONAL MUSEUM vol.82 



have preceded fixation, while the reverse is generally presumed to 

 have happened. (See Raymond, 1921, p. 347.) In the present case, 

 however, the coenenchym layer and the float provided the stratum 

 of fixation for the succeeding generations of zooids, thereby allow- 

 ing the development of budding. It seems quite logical that this 

 should have been a stage preceding that of discovery of and fixation 

 on the ocean bottom. (See the opposite view of Alexander Agassiz 

 below.) The gradual or accidental sinking of such colonial stocks to 

 a favorable bottom would have provided the transitional stage from 

 the floating colonial stock to the fixed one. 



C amptostroma is then to be regarded as an early hydrozoan that 

 had not yet progressed from the pelagic habit to the bottom (ben- 

 thonic) habit, and was still in the first stages of developing a chitin- 

 ous (or mucine ?) skeleton. 



I have before emphasized on the one hand the striking similarity 

 in outline and coarser sculpture (the ribs) of the disks here de- 

 scribed to the floats of the siphonophores, as represented by the re- 

 cent Porpita, the Devonian Paropsonema^ and possibly the Middle 

 Ordovician Dlscophyllutn; and, on the other hand, the close homol- 

 ogy of the details of the skeleton of the disk with those of the coe- 

 nenchym of the Tubulariae. It is of great importance that lead- 

 ing authorities on the Hydrozoa, as Kolliker, Louis Agassiz, Mc- 

 Crady, and Alexander Agassiz (1881, p. 10), have pointed out that 

 the relationship of the Velellidae and Porpitidae to the tubularian 

 hydroids is very close. Agassiz describes very fully how Porpita 

 might be derived from a Hydractinia, or Podocoryne " in which the 

 chitinous extension of the base of the coenosarc may perhaps be con- 

 sidered as the first indication of the formation of the float " {op. cit. 

 j)p. 10, 11). He goes even further in indicating the homologies in 

 the zooids of the Tubulariae and the Siphonophora mentioned and 

 {op. cit. p. 12) in pointing also to the close relationship of Porpita 

 to the Hydrocorallinae, the singular wiiite plate of the float, and 

 its peculiar structure reminding him of the porous structure of the 

 corailum of Sporadopora, Allopora, and Millepora, and finally even 

 in mentioning the Stromatoporae, which, if related to Millepora, 

 would carry back the hydrozoans to the Silurian. Summarizing 

 these observations, it seems safe for us to consider G amptostroiiia as 

 a tubularian hj'^drozoan with relations to the Siphonophora on one 

 hand and the Paleozoic Stromatoporoidea on the other, to the an- 

 cestors of all of which it may stand in close relation through its 

 generalized character. It is to be remembered that tubularian 

 Hydrozoa and Siphonophora apparently were already present in 

 Lower Cambrian time.^ 



*! have recently described (1931, p. 2) a Middle Cambrian hydrozoan {CJiaunograptus 

 scandens) of the campanularld type (order Calyptoblastca). 



