THE " PORCUPINE " AND " LIGHTNING " TUNICATA. 225 



and most of the specimens look like little rough bullets covered with sand (see 

 Plate XXXV. fig. 4, e. and f.). They feel quite hard, the test being rather firm. 

 The specimens collected vary from 2 mm. to 9 mm. in greatest diameter. 

 Most of them are small. In the majority, the apertures are not visible 

 externally, and it is impossible to distinguish the ends and sides without dissec- 

 tion. Ina few, however (see Plate XXXV. fig. 4, a and b.), the apertures are pro- 

 minent, terminating short conical projections from the anterior end of the body. 

 No lobes are visible, but when the test is removed the apertures are seen to be 

 distinctly cross- slit. 



The mantle does not adhere to the test, and consequently the body can 

 be readily shelled out. The musculature is well developed all over, and consists 

 of two distinct layers, the internal longitudinal, starting anteriorly in bundles 

 of fibres radiating from the apertures, and the external circular. Besides these, 

 there are also a few oblique and irregularly running bundles. 



The branchial sac appears variable. In small (young) specimens (see 

 Plate XXXV. fig. 6), the stigmata are short and rounded, and the transverse 

 vessels very wide; while in the larger specimen examined, the stigmata are 

 long (see Plate XXXV. fig. 5) and closely placed, and the transverse vessels all 

 very narrow. The internal longitudinal bars are wide and ribbon-like. In the 

 part of the sac of the large specimen examined (see Plate XXXV. fig. 5) there 

 were five bars on each of two folds next the endostyle, and only a single bar in 

 the interspace, while the two rows of meshes formed by this bar with the adjacent 

 folds had from five to six stigmata in each mesh. The series next to the endo- 

 style was wider, each mesh containing nine or ten stigmata. In the young speci- 

 men examined and figured (PI. XXXV. fig. 6) the first or dorsal fold (br. f. I.) 

 has seven bars, and is separated by a single row of meshes from the dorsal 

 lamina, and by four rows of meshes from the second fold — hence this interspace 

 has three bars. The second fold {br. f. II) has three bars, and is separated 

 from the third by three rows of meshes, hence this, the second interspace, has 

 two bars only. The third fold (br. f. III.) has five bars, and is separated 

 from the fourth by three rows of meshes, hence this third interspace has also 

 two bars. The fourth fold (br. f. IV.) has also five bars, and is separated from 

 the endostyle by two rows of stigmata, or an interspace with one bar. The 

 stigmata in this sac are all short and rounded, and placed far apart. There 

 are usually three or four in a mesh. 



The tentacles (Plate XXXV. fig. 6) are rather irregular. Three sizes are pre- 

 sent, but members of the third order are often absent, as seen near the endostyle 

 at the left hand end of the figure. The polycarps are fairly numerous. Some 

 are male, others female, and others hermaphrodite. The endocarps are rare. 

 The stomach is globular, and deeply sulcated. 



