VELELLA. 109 



pages I shall therefore borrow freely from the excellent memoirs of Kolliker and Vogt, 

 contained in the works which I have so often had occasion to quote. 



The body of Velella (fig. 1 ) presents a firm, horizontal, rhomboidal disc, convex upwards 

 or proximally, in the middle and fringed by a delicate membranous limb. The crest passes 

 diagonally from one angle of the rhomboid to the other, and, like the horizontal disc, presents 

 a firm central portion, and a soft marginal limb. It is triangular in shape (at any rate in most 

 species), the apex of the triangle being immediately over the centre of the horizontal disc. 



In its natural position the Velella floats on the surface of the sea, with its crest vertical 

 and exposed to the air,^ so as to act as a veritable sail ; and, consequently, very little is to be 

 seen of its appendages, which are all situated on the under surface of the horizontal disc ; and 

 indeed of its firm part, as none are attached to the membranous limb. 



The large central polypite presents a projecting, conical, often plaited, oral extremity, 

 which is susceptible of considerable dilatation. Its base is more particularly prominent m 

 the line of the crest, so that the polypite is not altogether symmetrical. On its base, and on 

 the under surface of all firm portions of the horizontal disc, are scattered innumerable minute 

 polypites, which perform the ofiSce of gonoblastidia, and slender simple tentacles, with slightly 

 enlarged extremities, spring in a single series from the line of junction between the firm part 

 and the membranous limb. 



The difi"erence between the "firm part" and the membranous "limb" in the horizontal 

 and vertical portions of the body of the Velella, is caused by the presence in the former of a 

 hard resisting body, which is commonly known as the shell of the Velella, and consists of a 

 horizontal and a vertical plate, which are perfectly continuous with one another. The notion 

 entertained by Lesson that this shell, the pneumatocyst, is composed of several pieces, has arisen, 

 as Kolliker has well pointed out, from a mistaken view of the nature of certain markings on 

 its surface. Thus there is a linear groove-like depression (fig. I w, iv), which traverses the upper 

 face of the horizontal portion of the pneumatocyst, nearly at right angles to the crest, and 

 rises upon this in the middle line until it reaches its apex. A slight ridge on the under sur- 

 face of the pneumatocyst answers to this groove, while a longitudinal depression, increasing in 

 depth from the margins to the centre, corresponds with the attachment of the crest. 



The horizontal division of the pneumatocyst (accidently inverted in fig. 2) consists of two 

 thin laminae, passing into one another at their free edges, and united by a number of con- 

 centric^ vertical septa, between which are corresponding chambers filled with air. All these 

 chambers communicate together by means of apertures in the septa. Of these each septum 

 presents two, placed at opposite points of its circumference, and all nearly in the middle line 

 of the pneumatocyst. Kolliker made the interesting discovery that many of the chambers 

 have an additional opening, by which they communicate directly with the exterior. These 

 apertures are situated in the proximal or upper wall of the chambers, along a line about 

 midway between that of the openings just described and that of the vertical plate of the 

 pneumatocyst. Of the thirteen apertures observed by Kolliker, six lay on one side of the 



' At least, I always observed it in this position. Delle Chiaje and Kolliker, on the other hand, 

 seem to have seen it with the crest turned downwards. 



' Vogt and others have affirmed the horizontal portion of the pneumatocyst to be formed of a 

 spirally coiled tube ; but this is an error. 



