100 WYVILLK THOMSON^ OX ASTERACANTHIOK VIOLACErS. 



growth^ Avitlieriug and disappearing when the permanent 

 external form and internal structure of the embr\o became 

 well defined. 



The appendage was attached to the dorsal surface of the 

 embryo in Echinaster sanguinolentus, and apparently toAvards 

 the oral aspect in Asteracanihion Mnlleri. Sars' impression 

 was that the cicatrix indicating the point of separation was 

 the madreporiform tubercle. Sars observed an opaque tubercle 

 in the centre between the four terminal suckers of the pedun- 

 cular appendage, but could detect no mouth-opening in this 

 position. 



Desor (^Proc. Boston Soc. of Nat. Hist./ February, 1848) 

 describes a mode of development slightly different, but on 

 precisely the same type, in an American Star-fish. In this 

 case the peduncle is simple, and depends excentrically from 

 the oral surface of the embryo. Desor regards the peduncle 

 as a vitelline sac, and believes it to be directly connected 

 with the digestive system, into whose general cavity its con- 

 tents are gradually absorbed. 



Agassiz (^ Lectures on Comparative Embryology,^ Boston, 

 1849) confirms Desor's observations, but gives no definite 

 opinion on the relations of the accessory appendage. 



Busch {' Beobachtungen fiber Anatomic und Entwickelung 

 einiger wirbellosen Seethiere,^ Berlin, 1851) describes the 

 development of Ecldnaster sepositus. The embryo of this 

 species closely resembles that of Ech. sanguinolentus, de- 

 scribed by Sars. Busch, however, figures the peduncle as 

 disappearing at the oral surface, and he describes a mouth 

 in the centre of the peduncle, between the four suckers. It 

 is unfortunate that Johannes Muller, the great authority 

 on echinoderra development, had no opportunity of obsers'- 

 ing any of this group of embryos alive, all his observations 

 having been made on swimming larvse taken with the towing- 

 net in the open sea; he examined, however, carefully, 

 specimeus sent to him in spirits, could detect no mouth- 

 orifice to the peduncle, and concluded that the hollow 

 suckers had no immediate connexion with the stomach, which 

 was developed as a distinct sac at some distance from their 

 point of attachment. 



According to Busch, the plan of development in Astera- 

 canchion glaciulis (L.) is somewhat diff'erent, associating 

 itself apparently with the very interesting type described by 

 Koren and Danielssen {' Fauna littoralis Norvegiae,' part ii, 

 Bergen, 1856) in Pter aster militaris (M. and T.), to which I 

 shall have to refer hereafter. 



Several writers, and particularly Dr. Carpenter, in his 



