66 ME. p. II. CAEPENTEK ON THE GENUS ACTDsOJMETEA. 



course, so as to give a leaf-like appearance to the rays of the non-pigmcnted intcrradial 

 star ou the ventral surface of the centrodorsal plate. 



In another specimen of the type, however (PI. VI. fig. 3), the sides of the narrow 

 basal grooves are almost parallel, and in every case meet at some distance within the 

 margin of tlie plate, while the intcrradial ridges residting from their union scarcely 

 extend at all beyond the angles of tlie external pentagon. 



Lastly, in the monstrosity represented in PL VI. fig. 10, the dorsal aspect of which 

 is seen in PI. II. fig. 8, both ridges and grooves are extremely indistinct, and in no case 

 reach the outer margin of the plate ; while the margin of the internal cavity is markedly 

 pentagonal in form, and not ten-sided nor circular, as is the case where there are five 

 distinct intcrradial elevations alternating with the five radial areas (PI. VI. figs. 3, 8, 15) ; 

 and it does not project inwards so far as to conceal all the openings (u) of the canals 

 leading to the marginal cirrhus-sockets, as is the case in the other two specimens of the 

 type (figs. 3, 8) and in var. 1 (fig. 15). 



In all these three specimens of the type the ventral surface of the centrodorsal plate 

 is not nearly so flattened as in var. 1, but rises very distinctly between its external and 

 its internal margins, Avhilc the radial areas are marked in the same way by various 

 indistinctly marked radiating ridges and furrows ; though as the floor of the central 

 cavity is also somewhat thicker, its depth is but little if at all greater. The same is the 

 case in the other tlirce varieties, in each of which the basal grooves differ slightly in 

 form from one another and from the type. In every case they are widest about the 

 middle of their length, as in one of the specimens of the type (PI. VI. fig. 8, b.g) ; this 

 is most marked in var. 3 (fig. 21), and least in var. 2 (fig. 17). They reach almost, if not 

 quite, to the margin of the plate, though the ridges formed by the union of their sides 

 extend but little if at all beyond it, except in var. 3 (figs. 20 & 21), in which two of the 

 angles of the external pentagon are marked by traces of the small processes (t) so 

 distinctly seen in var. 1 (figs. 14, 15). In this variety the course of the intcrradial 

 ridges is indistinctly visible on the dorsal surface of the plate (fig. 20), which is slightly 

 hollowed in the centre. The floor of the central cavity is, however, very thick and solid, 

 and its middle portion presents no trace whatever of any perforations for the canals of 

 preexisting cirrhi, though those proceeding to tlie three marginal cirrhus-sockets are just 

 visible under the projecting lip (fig. 21, u), which conceals several others. These indicate 

 that more cirrhi either have been or would have been developed had the animal lived 

 longer, their external openings having been obliterated in the former case (the more 

 probable one) and not yet formed in the latter. 



(§ 45) In Act. sol(t7'is (PI. V. fig. 2) the ventral surface of the plate-like centro- 

 dorsal piece is very nearly flat, as in Ant. rosacea {VI. IV. fig. 15) and \\\ Act. polymorpha, 

 var. 1 (PI. VI. fig. 15), rising but slightly from the circumference towards the centre, 

 and marked by five intcrradial elevations, along the top of each of which runs a long and 

 narrow basal groove {b.g). Its width is tolerably uniform from its base until near its 

 end, M'hcre its sides suddenly approach one another, and meet at a little distance within 

 the margin of the plate, where the ridge formed by their union also ceases without 

 extending outwards beyond the general surface of the plate. The same is the case in 



