VELELLA. 115 



On two of its sides, in a plane perpendicular to that of tlie crest, tliere was a double 

 crescentic mark, caused by a depression. The air did not completely distend the pneumatocyst, 

 but appeared to be divided into seven or eight lobes below, so that, at first sight, the organ 

 itself appeared to be lobed, but this was not really the case. It was, in fact, in the 

 smallest specimens a simple vesicle, about one twentieth of an inch in diameter, with strong 

 and thin walls, which, when it was burst and the air expelled, fell into sliarp folds. 



In the individual figured the commencement of the first chamber is indicated by a line, 

 concentric with the outhne of the pneumatocyst, which corresponds with the base of 

 the first septum. 



A Velella about twice the size of the foregoing (fig. 14) had a flatly campanulate 

 disc, with a nearly circular outline. The crest was oblong, raised to an obtuse point in the 

 middle, and somewhat narrowed at its attachment. The edges of the disc and crest had 

 a blue fringe, internal to which were short, radial lines, of a yellowish-red colour, arising 

 from an aggregation of corpuscles similar to those contained in the medusiform zooids. 

 The rest of the animal was colourless. 



The pyramidal pneumatocyst inclosed a considerable quantity of air, and presented 

 several concentric lines on its surface. The central polypite appeared somewhat four- 

 lobed at its base, and contained the half-digested remains of small crustaceans. A 

 single series of tentacles surrounded the polypite. Like those of the adult they were 

 slightly enlarged at their extremities, which were covered with thread-cells, and presented 

 a vacuolated a.xis marked by a dark line in its centre. They were attached in the re-entering 

 angle between the limb and the central polypite. Apparently proceeding from the outermost 

 margins of the latter were several small, ovate sacs, with two or more minute papillae, 

 containing thread-cells at their apex. The ectoderm and the ciliated endoderm could be plainly 

 distinguished in the thin walls of the central polypite, and these small sacs, containing 

 a cavity whose walls were likewise ciliated, and had precisely the same structure as those of 

 the central polypite, were, I believe, simple processes of the latter. 



The crest (fig. 12) was thin and membranous throughout, and divided by vertical septa 

 into a number of canals. The innermost of these passed vertically upwards, but the outer 

 were more or less curved in correspondence with the outline of the lateral margins of 

 the crest. At the margin of the crest it was easy to see that the outer wall of the marginal 

 canal was double, its ectoderm being lined by a distinct endoderm. 



A small quantity of brown, hepatic matter had accumulated under the pneumatocyst in 

 this individual. The disc was thicker and denser than the crest, but I did not work out its 

 structure in this specimen. 



In another of nearly the same size, or a little larger, the structure of the disc (fig. 1 1) was 

 essentially the same as in the first-described specimen, but a number of reddish granules had 

 accumulated in the marginal canal and in the peripheral extremities of the radial canals. 



In a still larger individual the radial canals had become much narrower in proportion 

 to their length (fig. 13). At their central ends their walls were simple and straight, 

 and separated only by a thick, clear, transparent substance, but, externally, they were 

 sharply folded, or drawn out into short, lateral sacculi ; and, from the ends of these, 

 ramified processes, as a sort of connective tissue, extended into the clear substance which 

 separates the canals. At their peripheral ends, the canals, instead of being simply divided 



