76 THE VOYAGE OF H.M.S. CHALLENGER. 



supporting lamella and the layer of nerve fibres of the ectoderm. Jt is necessary to 

 employ a staining fluid (picro-carmine), which impregnates the supporting substance 

 strongly, in order to distinguish a framework of connective tissue between the muscular 

 fibres ; and as this gives rise to extremely fine-walled meshes, it divides the muscular 

 fibres into mesodermal bundles of fibrillar. Numerous supporting layers run out from the 

 surfaces of the supporting lamellae, and these ramify and anastomose with one another. 

 The anastomoses are wanting towards the ectoderm, so that the meshes open towards the 

 layer of nerve fibres ; the muscles are consequently partly mesodermal, partly ectodermal. 



In Calliactis p>arasitica the boundary line between the mesodermal muscles and the 

 ectoderm is also indistinct, but the bulk of the former is much smaller, so that sparse 

 bundles only are enclosed in an abundant fundamental substance. 



The tentacles of Calliactis pxylypus are long, slender, and end in a fine point. I made 

 out about seventy in a twelfth part of the animal, so that altogether they amount in 

 number to several hundreds, which decrease in size from within outwards, and are 

 arranged in about ten circles. The first and second circles, beginning at the inside, each 

 contain twelve tentacles placed somewhat apart, the third twenty-four tentacles, the 

 fourth forty-eight, and so on. 



If we take the well-developed septa only into consideration, there are altogether 



our cycles or forty-eight pairs, the first six pairs of which are perfect. The following 



six pairs are imperfect and sterile like the first six, so that the reproductive organs are 



confined to the septa of the third and fourth orders. The specimen examined by me was 



a male, and contained ripe testes. 



Cereus, Oken. 



Sagartidse, with numerous tentacles and circular oral disk, without cinclides which 

 can be anatomically demonstrated ; wall rough, and covered with knobs. 



Milne-Edwards (Hist, des Corall., torn. i. p. 263) included all the more typical represen- 

 tatives of his " Actinines verruqueuses " in Oken's genus Cereus. After it had been 

 shown that acontia existed in Cereus bellis, which had been taken by Oken as the typical 

 representative of the genus (Lehrbuch d. Naturgeschichte, Th. III. Abth. 1, p. 349, 1815), 

 Verrill limited the name to forms of the family Sagartidse. I agree with Verrill on this 

 point, but wish to attach more importance in the diagnosis to the papdlose nature of the 

 wall, in order to establish a sharp distinction between this genus and Sagartia. I have 

 therefore altered the description of the wall, which runs thus in Verrill : " upper part 

 with small, inconspicuous contractile suckers ; walls nearly smooth." 



Cereus spinosus, n. sp. (PI. I. figs. 3-5; PI. VI. fig. 1 ; PL VIII. fig. 6; PL XII. fig. 10). 



Papillae of the wall unequal in size, with a tendency to arrangement in transverse and 



longitudinal rows ; each papilla runs out into a fine point, which is placed on a hemi- 



