ix ENTEROPNEUSTA ONTOGENY 589 



In Tornaria Agassizi, it was observed that the pores of the second pair of gills 

 arise earlier than those of the first. 



In other cases, the new pairs of gills always form behind the last produced. 

 During the process, the growing oesophagus, the posterior part of which is now the 

 branchial intestine, passes from a vertical to a horizontal position. 



Even in the adult animal, new gill-pouches are constantly being formed at the 

 posterior end of the branchial region. The course of development is always the 

 same as that of the first gills in the larva. 



The collar canals, with their pores, develop shortly before the rudiments of the 

 second pair of gills arise, most probably as outgrowths of the anterior wall of the 

 first gill-pouch, near its external pore. These outgrowths run forwards towards the 

 collar ci.elom. 



The first rudiment of the proboscidal diverticulum of the buccal cavity has 

 been observed in Tornaria Agassizi, as a small bulging of the anterior wall of the 

 larval oesophagus directed anteriorly, and lying immediately above the mouth. 



The proboscidal ccelom. The rudiment of the proboscidal ccelom has not as 

 yet been observed with as much certainty as could be desired. According to one 

 observer, it arises as an outgrowth of the intestine at the boundary between the 

 oesophagus and the stomach. 



In the youngest stages which have been closely observed, the proboscidal ccelom 

 (the so-called water sac of the larva) is an almost cylindrical tube lined with 

 tesselated epithelium, the inner half of which is slightly widened. This tube 

 becomes attached by its inner end to the anterior wall of the oesophagus, near 

 the point at which the latter opens into the stomach. Here it sends two processes 

 (the reins) to right and left on to the lateral walls of the oesophagus, on which 

 it appears to ride. From the oesophagus, the tube, traversing the blastoccel, 

 ascends almost vertically. Shortly before it reaches the dorsal ectodermal wall, it 

 is continued into a short internally ciliated tube, the epithelium of which suddenly 

 changes its character. This short tube, lined with cylindrical epithelium, is the 

 rudiment of the proboscidal canal, and the pore to which it leads is the proboscis 

 pore. 



The proboscidal coelom is further attached to the apical plate by means of a 

 strand in the manner illustrated in Fig. 466 ; the strand consists of contractile 

 fibres surrounded by a nucleated envelope, a continuation of the wall of the water 

 sac. The contractile fibres of this strand are continued on to the tips or reins of 

 the water sac, which grasp the oesophagus between them. 



The further fate of the water sac, briefly stated, is as follows. It swells up, 

 and soon changes from a tube into a vesicle, the greater part of the wall of which, 

 during metamorphosis, becomes applied, as parietal wall of the proboscidal ccelom, 

 to the body epithelium of the anterior or proboscidal section of the body, the 

 muscular apical strand becoming shorter and shorter, and finally altogether 

 disappearing. 



The manner in Avhich the epithelial walls of the water sac become differentiated 

 into the proboscidal musculature cannot here be described. 



" Heart vesicle." Authorities differ as to the first rudiments of this organ. In 

 the youngest stage investigated by the most recent observers, it is a small cellular 

 structure with an internal cavity lying so close to the ectoderm that it may be 

 either ectodermal or mesenchymatous. It appears to the right, in front of and 

 near the proboscis pore. The body becomes a hollow vesicle, leaves the ecto- 

 derm, sinks below the surface, and becomes applied to the right side of the water 

 sac. Transverse muscle fibres develop on its ventral wall. The water sac then forms 

 posterior outgrowths, which grow round the "heart vesicle'' on its right dorsal 

 and ventral sides. The two dorsal posterior and the ventral posterior sections 



