80 ME. P. H. CAEPEXTER OX THE GENUS ACTIXOMETEA. 



schnittc dci* Lcibesliohlc in Schcibc und Armcn stclien mitcinander paarwcisc in eng-crer 

 Beziehung, indcm sich dicaxiale Lcibesliolc fortsetzt in die Ventralcaniile [subtentacular] 

 der Armc und Pinnula3, die intcrviscerale in die Dorsalcanjile [coeliac] und die circum- 

 viscerale in die Genitalcanale." 



I cannot corroborate this statement except with regard to the axial body-cavity, the 

 connexion of whicli with the subtentacular canals of the arms was first shown by 

 Dr. Carpenter. 



The ventral portion of the circumvisccral body-cavity, viz. the limited and much 

 divided space between the parietal and visceral layers of the ventral peritoneum, certainly 

 does stand in direct connexion with the genital canals of the arms ; but its dorsal 

 portion, viz. the space between the visceral mass and the skeleton of the calyx, is, as 

 already mentioned, far more du'ectly a continuation of the coeliac than of the genital 

 canals. The former gradually increase in size as they approach the disk, becoming very 

 large in the second and first arm-segments, and traversed by numerous connective-tissue 

 bauds, which are directly continuous with those of the circumvisccral space ; while the 

 genital canal remains relatively small, and is nothing more than a space in the horizontal 

 septum separating the subtentacular and coeliac canals. 



The beautiful Investigations of Gottc ' have shown that the primitive coelom of the 

 pentacrinoid larva of Antedo)i consists of two parts : (1) an oral or ventral one, de- 

 veloped from the left peritoneal diverticulum of the primitive alimentary canal ; and 

 (2) a dorsal one, which sends a prolongation backwards into the stem, and is developed 

 from the corres^jondiug right peritoneal diverticulum. These divisions of the primitive 

 ccelom had been previously described by Dr. Carpenter ^ Metschnikoff '^, and Greeff *, 

 to all of Avhom, however, their origin was unknown. The last observer regarded the 

 ventral division as " den vom Wassergefasssvstem und der hinteren Leibeshohle "eschie- 

 deuen ursjiriingliclieu Blutsinus ;" for he supposed it to be continuous with the cavity 

 of the axial prolongation, which he called the " dorsovcntral Gefiissstrang." Dr. Cai'- 

 penter has found, however, that this structure breaks up into five branches, one of which 

 goes to each of the primitive rays, and devclopes into the so-called " genital rachis " of 

 the arms, while the oral ca3lom of the pentacrinoid larva (the "Bloodsinus " of Greeff), 

 sends off an extension into each of the arms as its subtentacular canal. In the direction 

 of tlie radii it forms, of course, the subtentacular canals of the disk ; but elsewhere, or 

 interradially, it becomes gradually limited by the enlargement of the visceral mass, and 

 by the formation of adhesions between its upper surface (visceral layer of the peritoneum) 

 and the parietal layer lining the under surface of the ventral pcrisome ; so that the 

 ventral portion of the circumvisccral coelom enclosed between these layers, to which the 

 primitive oral coelom gives rise, becomes very much reduced in extent. "We do not yet 

 know the precise origin of the genital canals of the arms ; but it seems most probable 

 that, like the ventral portion of the circumvisccral coelom with which they arc connected 

 in the disk, they are developed out of the lower or dorsal half of the extension into the 



' Op. cii. p. 591, Taf. xxri. fig. 19. ' Phil. Trans, he. cit. p. 728 ; Proc. E. S. no. IGG, p. L'28. 



' " Beitr. z. Entwickclungsgesch. ciniger niederen Thiere,'' BuU. de TAcad. Imp. des Sciences de St. Pctersb. loin. xv. 

 1871, pp. 502-509. ■• Marburg Sitzungsberichte, 1870, Xo. 5, p. 89. 



