THE "TRITON" TUNTCATA. 103 



shape, hence it is probable that the abnormal form is clue either to the 

 animal not having been killed suddenly enough or to imperfect preservation. 

 All of the " Triton " specimens, with the exception of the five small ones 

 mentioned in the list on page 93, are between 6 mm. and 12 mm. in length, 

 and most of them measure 1 cm. This size is apparently much greater than 

 that of Mediterranean specimens, as Grobben speaks of his as being about 

 2 o mm., while Keferstein and Ehlers figure one 3 mm. in length. 



Most of the specimens are in ordinary rectified spirit, while a few have been 

 treated in each of the following methods : — 



1. Preserved in absolute alcohol. 



2. Put first into chromic acid solution and then into absolute alcohol. 



3. Preserved in a saturated solution of picric acid. 



4. Put first into solution of osmic acid and then into absolute alcohol. 



5. Put first into solution of picric acid and then into absolute alcohol. 



6. Preserved in solution of chromic acid. 



These specimens were all in excellent condition for examination, and the 

 different methods appear to give almost equally good results. Perhaps the best 

 preparations for most histological points were obtained from the specimens 

 preserved in chromic acid by thoroughly washing in alcohol, staining in picro- 

 carmine, and mounting in Farra.nt's solution ; while for some few special 

 points the specimens preserved in osmic acid solution and absolute alcohol 

 excelled. 



The test is almost absent, being represented merely by a delicate structure- 

 less layer over the ectoderm, which covers the surface of the mantle. The 

 mantle contains the muscular bands or hoops, which, in this form, are eight 

 in number (m 1 to m 8 in the figs.). The first and last of these bands form 

 sphincters for the apertures, and usually appear to terminate the body anteriorly 

 and posteriorly, as shown in Plate XVIII. fig. 4, the delicate denticulated 

 margins of the branchial and atrial apertures being almost invariably turned 

 in or directed across the opening. This denticulated margin was turned out 

 in the chromic acid specimens examined, and was more perfectly preserved 

 than in any of the others. It is divided into twelve lobes around the branchial 

 aperture and ten around the atrial. The muscle bands are composed of very 

 long non-striped fibres, closely and regularly placed, as shown in Plate XVIII. 

 fig. 6. Sometimes, as in fig. 5 (from a picric acid specimen), the fibres are 

 thrown into undulations. 



The wide branchial aperture leads into the branchial siphon, which, as there 

 is no diaphragm and no circlet of tentacles, may be considered as extending back 

 to the peripharyngeal band. This band, in all the specimens which I have 

 examined, runs in most of its course between the 2nd and 3rd muscle bands, or 

 in the 2nd intermuscular space (PL XVIII. fig. 11, p.p), and marks the anterior 



