238 M. Sars on the Development of Star-fishes. 



forwards, in becoming fixed at a later period, proves itself to be 

 in reality the posterior. 



In regard to those species of Star-fish that are provided with 

 several madreporoid plates, no theory can at present be hazarded 

 with any degree of probability. Perhaps they possess at an early 

 period several separate organs of attachment. 



In conclusion, I would insist upon the naturalist being very 

 guarded in his attempts to generalize, for it is probable that when 

 the relations of development between other genera and species of 

 Star-fishes come to be more closely investigated, greater differ- 

 ences will be detected than could have been a priori surmised. 

 Thus Asteracanthion rubens differs distinctly from Echinaster san- 

 guinolentus, whose genital openings must be placed upon the 

 ventral side, inasmuch as, according to Muller and Troschel, these 

 very apertures are found upon the dorsal region in the Astera- 

 canthion rubens, on which account its eggs probably fall into the 

 sea and are left to take care of themselves. This is perhaps the 

 reason why, in spite of my industrious researches at the most dif- 

 ferent times of the year, I have never found the young, or even 

 any indication of incubating instinct in this species. Another 

 species of the same genus, Asteracanthion Mulleri, agrees, as we 

 have seen, with Echinaster sanguinolentus. The development of 

 some other Star-fishes appears to be still more anomalous. Thus 

 the animal, formerly* called by me Bipinnaria asterigera, I con- 

 jecture, from more recent investigations (to be published upon 

 some future occasion), to be only a Star-fish provided during its 

 development with a large natatorial apparatus. 



I cannot here refrain from observing, that the development of 

 the Star-fishes, so far as we are acquainted with it, exhibits im- 

 portant departures from that of the other Radiated animals, the 

 Polyps and Acalephse, a greater number at least of which are di- 

 stinguished by the peculiar form of the incubating organ, or 

 marsupium. The Star-fishes are developed without any such 

 change in the generative process from the state of an ovum to 

 the peculiar type of their group, and in this respect agree with 

 the Articulata and Vertebrata, to which they form the first ap- 

 proximative step, both by their peculiar pointed calcareous ske- 

 leton, and in the remarkable instinct with which they nurse their 

 young.] 



EXPLANATION OF PLATE III. 



Fig. 1. An Echinaster sanguinolentus, natural size, seen from the ventral 

 aspect, with the incubating cavity half open, disclosing within the 

 bright red-coloured young. 



Fig. 2. The same seen in profile, resting with the cavity completely closed ; 

 a, the madreporoid plate. 



* Beskrivelser og Jagttagelser, etc. p. 37. tab. 15. fig. 40. 



