324 Prof. E. W. MacBride. The Artificial Production of 



anatomy of the early larva is of the same type. In all three groups the 

 larva possesses a simple alimentary canal, consisting of a conical oesophagus 

 opening by a wide mouth, a globular stomach, and a sac-like intestine 

 opening by a narrow anus and directed forwards, so that the whole 

 alimentary canal has the form of a U. On each side of the oesophagus 

 a flattened coelomic sac is situated ; of these, the left sends up a vertical 

 outgrowth termed the pore-canal, which fuses with the dorsal ectoderm, 

 and opens to the' exterior by a pore called the madreporic pore. Each 

 coelomic sac subsequently grows backwards, so that its posterior portion lies 

 beside the stomach, and this portion later becomes separated by a constriction 

 from the rest. Consequently, each sac becomes divided into an anterior 

 and a posterior coelom. 



In all three orders, when this stage has been reached, a vesicle is budded 

 oflf from the posterior end of the anterior coelom on the left side, which is 

 termed the hydroCCEle. This vesicle dominates the whole later develop- 

 ment of the larva, and eventually becomes transformed into the water-, 

 vascular system of the adult. It gives off lobes, or primary tentacles, 

 which become converted into the radial vessels of that system, and it 

 becomes bent into the form of a hoop, which is then converted into a ring 

 by the meeting of the two ends, and this ring is the water-vascular ring- 

 canal. The axis of adult symmetry, as all know, passes through the centre 

 of this ring. At metamorphosis, portions of the larval body are cast off, 

 consisting mainly of the ciliated band, with its various outgrowths or 

 " arms," which constitute the organs of locomotion of the larva. The left 

 side of the larva becomes converted into the ventral or actinal side of the 

 adult which bears the tube-feet, whereas the right side of the larva becomes 

 changed into the dorsal or abactinal side of the adult, developing spines and 

 pedicellariae as the case may be. 



It is obvious that, if not only the left, but also the right, anterior coelom, 

 were to bud off a hydrocoele, the bilateral symmetry of the larva would be 

 retained, and the whole subsequent course of the development would be 

 profoundly modified. 



Johannes Muller figured, without describing it, an Ophiurid larva, with 

 a five-lobed hydrocoele on both sides of the oesophagus (16). Later, 

 Metschnikoff (15), who investigated the development of the viviparous 

 Ophiurid AmpMura squamata, asserted that the rudiment of a hydrocoele 

 was normally formed on each side of the embryo, but that only the left 

 rudiment developed. 



In 1895, when engaged in the investigation of the development of the 

 Asterid Asterina gibbosa, I encountered several specimens with_ two 



