356 E. W. MACBRIDE. 



{ect.y fig. 5). This is caused,, as is shown by fig. G, by an 

 enlargement of some of the cells, whose protoplasm becomes 

 at the same time clear and glassy. These enlarged cells sink 

 inwards, and are covered by the adjoining unaltered ectoderm 

 cells. A first trace of this peculiar change can be noted even 

 in younger larvee, such as that shown in fig. 4. 



In a later stage (fig. 7) the atrial ridges have coalesced 

 to floor in the atrial cavity. The metapleural coelom on each 

 side is represented by a solid mass of cells, and some of these 

 cells are already being drawn out transversely. This is an 

 indication of their approaching modification into the muscles 

 of the floor of the atrial cavity. A few fibres are ali eady dis- 

 tinguishable by their refractive power (t. muse). In the right 

 atrial ridge a cavity is seen with indications of nuclei in its 

 v/alls. This space I at first supposed to be identical with the 

 metapleural coelom of the earlier larva. Lankester and Willey 

 call it a lymph canal, and suppose it to be derived from the 

 hollowing out of the ectodermic thickening seen in an earlier 

 stage. This latter view I now believe to be correct. In the 

 series from which fig. 7 is taken, one can distinctly see the 

 metapleural coelom in the extreme anterior portion of the 

 rio-ht atrial ridge. As one follows the series backwards, the 

 metapleural coelom changes into a solid plug of cells, and is 

 pushed to the one side by a mass of watery cells lying under 

 the external ectoderm, in which further back the cavity 

 appears. From this observation I have no doubt that 

 Lankester and Willey are right in assigning to the meta- 

 pleural canals a " pseudocoelic " origin. 



The question now is whether the rudimentary atrial cavity 

 has been formed by an invagination of the ventral ectoderm, 

 or whether it has been walled in by the farther downgrowth 

 of the atrial ridges. If figs. 5 and 7 are carefully compared 

 with one another, it will be seen that the latter view is the 

 only tenable one. These figures represent sections taken 

 from corresponding parts of two larvas, one with free ridges, 

 the other with a rudimentary atrial cavity. The lymph canal 

 in the older larva has been carried further downwards than 



