124 Prof. P. M. Duncan on the 



ingj and they are not simple bags with a contractile coat. 

 The bag is composed of very delicate, close, exceedingly 

 narrow circular fibres, which in the main are vertical, although 

 they depart from that direction in the gibbous parts. The 

 accessory structures are irregular transverse rows of small 

 elongate pigmented spots, which, in sections or oblique views 

 of the ampullffi, are seen to be the outer terminations of fibres 

 resembling the usual unstriped fibres of small spines, which 

 pass inwards to varying depths in the hollow of the ampullffi, 

 and are attached to reticulations of connective tissue (fig. 22). 

 These fibres exist in Psammechinus and are very suggestive 

 of less definite structures, which may be noticed in the abac- 

 tinal branchial tentacles of Calopleurus. These spots give 

 a very marked aspect to the outside of the ampullae, and their 

 inner prolongations probably prevent too sudden expansion 

 of the calibre of the ampulla, and may also assist in the con- 

 traction required for the infilling of the tentacle. A very 

 delicate epithelium and endothelium* occur. 



It is interesting to notice the great size of the ampuUai in 

 relation to the dimensions of the bases of their tentacles at a 

 slight distance from the apical system, and to observe that 

 the bags reach quite across the whole breadth of the arabu- 

 lacral plates. 



It a2ypears from the examination of the bases of the tentacles, 

 that the so-called geminous or reverse condition of the jjores of 

 the 2>tti^'S is not of the least jphysiological importance, and 

 whether there is a groove externally on the test between the 

 pores or whether there is a gramde or nodule between them is 

 of no significance in a natural system of classification. 



In the specimens which gave the above results the ampullae 

 were tense and tumid, but in some others, in which they were 

 seen to be empty and flaccid, there was some departure from 

 the exact shape, and moreover the internal fibrous arrange- 

 ment was not visible. This was doubtless due to the con- 

 siderable time during which the test had been in impure 

 alcohol. The two openings for the tentacular canals were, 

 however, very visible, and there was no vestige of any val- 

 vular structure. But a fortunate section of the tentacle of 

 one of the ampulla3 showed that the inner membrane is con- 

 tinuous through the pores with that of the ampulla. 



The Tentacles. (PI. XL figs. 26, 27.) 



These well-developed structures of Temnopleurus are similar 

 throughout the ambulacra, and the genus is therefore homoeo- 

 * See also Hamaiui, o/:). cit. pi. xiv. fig. 3. 



