116 THE OCEANIC HYDROZOA. 



by a short septum, separated into two widely sacculated branches, full of brown granules. 

 The canals were much nearer the inferior than the superior surface of the disc. 



VeleUm one third of an inch long had the disc flat and oblong, and the crest oblique. 



Incomplete as these observations are, they seem to me to prove sufficiently that, 

 in its early stages, Velella departs in no respect from the typical structure of the Physo- 

 phorida, and that the membranous limbs of both the crest and disc must be regarded 

 as vertical and horizontal processes of that part of the hydrosoma which represents the pneuma- 

 tophore of other Physophorida, anA whose cavity has become subdivided by vertical partitions. 

 The existence of these last, however, is nothing new, for if the pneumatophore of a Phjsophora 

 or of an Agalnia were flattened into a disc or pulled out into a crest, the septa which retain 

 the pneumatocyst in its place would s.ssume nearly the same position and relations as the 

 partitions between the canals in the young Velellce. 



The hepatic mass does not at first exist, but appears to be developed in contiguity with 

 the under surface of the pneumatocyst, like the cellular processes attached to this organ 

 in Phizophysa; and I imagine that the canals which eventually traverse it are mere sub- 

 divisions of the somatic cavity,' produced by the encroachment of the lobes of this organ, and 

 their mutual anastomosis. 



The pneumatic filaments remain as something peculiar to the Velellidan family. 



The Velella on which my observations were made were taken on various occasions in 

 the Pacific and Atlantic Oceans. Lesson's 'Acalephes' was the only work which I could 

 consult, and those who are acquainted with that singular though useful miscellany, will under- 

 stand why I soon gave up the attempt to identify the Velella I met with, with any of the 

 sixteen species which are enumerated therein. I can only say that some were right-handed, 

 and some left-handed, and that Eschscholz's Rafaria are obviously nothing more than 

 such young Velella as I have described above. 



M. Vogt describes a great peculiarity in the mode of growth of the Velella. He affirms 

 that at a particular point of the cii-cumference of the disc, the series of tentacles and of 

 gonoblastidia always appear to be interrupted, overlapping one another as it were, like the 

 ends of a broken ring. At this point, the young tentacles and the gonoblastidia are developed, 

 and an aggregation of their buds is consequently always found here. The buds of the 

 tentacles are, at first, cseca placed on the circular vessel of the limb, and " composed of 

 two layers — an epidermal, blue layer, and another more transparent layer, in which the 

 internal cavity of the bud is excavated." M. Vogt has obviously seen the ectoderm and 

 cndoderm. In the course of development the former becomes covered with thread-cells ; 

 the latter, increasing in thickness, " shows itself composed of muscular fibres." That is, 

 as I believe, it becomes vacuolated, the intermediate substance between the vacuoles being 

 the so-called muscular fibres. 



Kolliker says nothing about the existence of a special " centre of gemmation," such 

 as that described by Vogt ; and Leuckart denies its existence, affirming that the deve- 

 lopment of the new organs takes place among the old ones in all parts indifferently. 

 ('Z. N. K.,'p. 118.) 



' The homology of these canals with the somatic cavity of other Hydrozoa is well pointed out by 

 Vogt, p. 35. 



