390 DR ARTHUR T. MASTERMAN ON THE 



view of homology, and a large amount of labour and time have been devoted to 

 corroborating the process. It has been confirmed in vertical and horizontal series, and, 

 however disappointing a result from the point of view of generalisation, it must be 

 accepted as the true course of events. 



The horizontal series of figs. 63 to 70 illustrates a slightly later stage of D/E. In it 

 may be noted the five elements of the hydroccele — the pore-canal, the right lateral ccelom, 

 the central ccelom with constricted aperture, and the completely fused right and left 

 posterior cceloms. The series is chiefly useful for comparison with the next series 

 through stage E. The stone-canal evidently (fig. 63) commences opposite the inter- 

 radius between 4 and 5. The right lateral ccelom is closely comparable to the left, and 

 does not extend quite so far as the latter. The mesenteron (fig. 66) is still perfectly 

 oval in shape, with an opening at each end into the ccelom. 



We are now in a position to follow the structures in stage E, as shown in the 

 horizontal series (larval orientation) of figs. 79 to 86. In fig. 63 the dorsal horn of the 

 left posterior ccelom is closely pressed against the posterior wall of the pre-oral ccelom. 



In fig. 79 the contiguous walls have broken down, and the two cavities remain 

 confluent for some time. In fig. 80 the two dorsal radii (4 and 5) of the hydroccele have 

 become sharply defined, and the pore-canal now opens freely to the exterior on the 

 right side of the larva. Slight traces of the degenerated hypenchyme fill its lumen. 

 Slightly below and anterior to this (fig. 81) is the central ccelom, still communicating by 

 a small fissure with the anterior ccelom. Posteriorly to the pore-canal, on the right side, 

 can be seen the right posterior ccelom, which in fig. 82 communicates still by a wide 

 aperture with the anterior ccelom below the pore canal. 



In (fig. 82) the ventral horn of the left posterior ccelom comes into view on the 

 right side anteriorly, and the mesenteron is seen to still open into the posterior ccelom 

 by a small aperture. Between this section and that in fig. 83 the anterior end of 

 the mesenteron should, judging by D/E, open into the anterior ccelom, but such is not 

 the case. The anterior end constricts till the opening is completely closed, and all 

 connection between the cavities of anterior ccelom and of mesenteron is lost. At the 

 same time the mesenteron, which in D/E was somewhat sausage-shaped (fig. 66), bends 

 into a V dorsally, bringing the anterior and posterior ends close together dorsally (both 

 are cut across in fig. 83). The right side of the mesenteron then becomes flattened 

 (figs. 84 and 85), giving the mesenteron a semicircular outline in section. Covering 

 the whole of its right surface is the right lateral ccelom, which thus also appears semi- 

 circular in section, though hemispherical in shape. In fig. 83 we again recognise the 

 anterior process of the left posterior ccelom, with its process extending from right to left 

 immediately ventral to the pre-oral ccelom. In fig. 85 this process reaches almost to 

 the left side of the larva. Figs. 86 and 87 show the progress in axial symmetry. 

 Not only have 1 and 2 radii of the hydroccele advanced considerably, but the left 

 posterior ccelom has pushed out two radial extensions lying over their respective 

 hydroccelic elements. 





