MKDUS.K. 15 



ami ileutlritic in shape, forming a kiml of ornamental liordcr to the radial and circular 

 canals. Hacckel calls these diverticula " leberartigen Canal-Drusen," but there is no 

 evidence that they function as glands. The diverticula of the ra<lial canals in the 

 early stage of ('atdldema wildoui arc simply sh<jrt lateral outgrowths without any 

 branching. In the a<lult, though some of the diverticula are short and simple, most 

 of them are more or less Ijranched. The mode of branching is, however, very variable, 

 and .scarcely two diverticula are alike. It i.s important to notice that the diverticula 

 are long and that some nearly meet those from the adjacent radial canals ; there is 

 uo definite arrangement of the diverticula upon the sides of the canals. They are 

 s<jmetiraes in opp<Jsite pairs, sometimes alternate, and apparently develop wherever 

 there is a sufficient space. In some of the other species of Catahlema the circular 

 canal has diverticula upon its upper margin, similar to those upon the radial canals. 

 In Catablema tceldoni the diverticula of the circular canal have disappeared, and their 

 former pre.sence is just indicated by a few minute vestigial outgrowths. 



The most interesting feature oi this Medusa is the presence of filaments (Plate I., 

 fig. 4) along the inner side of all the tentacles. The filaments are closely packed 

 together, forming a kind of frill which extends along nearly the whole length of the 

 tentacle, being absent from the basal portion oidy. The crowding together of the 

 filaments, so as to form thick dense masses, depends entirely upon the contraction 

 of the tentacle. In a semi-contracted tentacle the filaments are about four deep, 

 transversely, and when the tentacle is closely contracted they are denser still. In a 

 fully-expanded tentacle the filaments are probably arranged in a single or doulile row, 

 and then they should somewhat resemble the ai^pearance of the filaments on the 

 tentacle of Ctenaria cteiiophora. In .some of the specimens, owing to great shrink.ige, 

 the tentacles have the appearance of long narrow ribbons, with one edge lined with 

 filament.s. 



The state of preservation of the specimens is not good for minute hist«jlogical 

 details, but is sutiicieutly so to show thatthe filaments are composed of ecto<lerm cells. 

 The filaments are solid, and have a central strand of mesoglaca. They are capable 

 of a certain amount of e.vpansion and contraction. There is no visible evidence that 

 the endoderm of the tentacle enters the filament. Sections through the tip of a 

 filament show that it contains numerous very minute nematocysts. 



The tentacles are frequently very long ; .some measure 60 Ui 80 mm. in length, 

 and are V)y no means fully expanded. They arc hollow and well supplied witli 

 ectodermal mu.scle fibres. 



The basal bulbs of the tentacles are laterally compres.sed and curve over the thick 

 margin of the umbrella. There is not the slightest indit^tion of a pigment spot on 

 the ba.sal bulbs, nor of any other kin<l of .sense organ. Between every tw<j tentacles 

 there is a small marginal bulb, which is probably capable of developing a tentacle 

 when anf»ther is needed. 



One specimen has nine long tentacles, with a siu.ill bulb in between every two 



