1879.] the Comatnlae of the " Challenger" Expedition. 391 



or three faint indications of sockets, very much as in the stellate 

 centrodorsal of Phanogenia. 



Stage 4. The adult condition succeeds to this. The centrodorsal is a 

 simple pentagonal plate 3'5 — 4 millims. in diameter, and situated 

 entirely within, and on the same level as, the radial pentagon. 



This series of changes in the centrodorsal does not proceed any 

 farther in Act. Jukesii, but in other species it may continue still farther, 

 or on the other hand close sooner. Thus the centrodorsal of a 

 gigantic Actinometra in Professor Semper's Philippine collection is in 

 stage 2 ; while those of two other " Challenger " species, from the 

 same locality and equally large, exhibit the next stage of meta- 

 morphosis (5). 



Stage 5. In No. 37 (of my list) the centrodorsal is a pentagonal 

 disk without a trace of cirrhus sockets. It is slightly beloiv the level 

 of the radials, and is only in contact with them by its inter-radial 

 angles, its sides being separated from their inner margins by linear 

 clefts. 



No. 49 is in the same condition, and so are two others of Semper's 

 Philippine species, except that one or two minute cirrhus stumps still 

 remain on the centrodorsal. Another large " Challenger" species, 

 also from the Philippines, is represented by four specimens. All of 

 these show the gradual obliteration of the cirrhus sockets and the 

 lowering of the centrodorsal to the level of the radial pentagon, or 

 even below it, together with the presence of clefts at its sides. These 

 may occasionally appear before the loss of the small cirrhus stumps, 

 as is the case in PTianogenia. 



Stage 6. In No. 6, another large Philippine species, the clefts are 

 somewhat wider but very shallow. They are deeper in No. 30, but 

 there is no trace of cirrhi on the stellate centrodorsal. Except in this 

 point (a very variable one, as seen above) this seems to be about the 

 condition of Phanogenia. 



Stage 7. The last stage is reached in another of Semper's speci- 

 mens which was purchased from the (xodeffroy Museum, and appears 

 to me to agree very closely with Lutken's Act. stellata. The centro- 

 dorsal is star-shaped, having a flat centre and five rays, the length of 

 which is about one-third the diameter of the centre. The points of 

 the rays abut on the radial pentagon at the synostoses of every two 

 contiguous radials and are therefore inter-radial. The re-entering 

 angles of the star are occupied by five clefts, each of which is some- 

 what planoconvex in shape. It is bounded centrally by the centro- . 

 dorsal plate, laterally by two of its rays, and peripherally by the inner 

 margin of a radial. These openings are large enough to admit the 

 point of a good sized needle for a short distance. 



The causes which lead to such remarkable changes in the appearance 

 and relations of the centrodorsal piece, are I think, partly to be found 



