446 T. J. EVANB 



body in structure and function. These, while retaining their 

 connecting processes and their position as lining-cells of the 

 ceratal blood-spaces (fig. 5, h.s.), are at times among the 

 largest cells in the body and exhibit remarkable secretory 

 activities during the period of digestion of a meal in the neigh- 

 bouring gut diverticula. At the same time they increase greatly 

 in size till, finally, their identity as cells of the connective tissue 

 is obscured, and only a thin envelope continuous with the 

 processes (fig. 6, e.) is free from accumulated secretion staining 

 deeply with basic dyes. Simultaneously with the deposition of 

 stainable material in the cytoplasm, a clear non-staining 

 sphere (fig. 6, n.&.) grows within the nucleolus, which in fixed 

 tissue is so hard as to be frequently displaced or torn out by 

 the microtome knife. In specimens with empty stomachs 

 these cells are found in various stages of reduction in size, 

 an early stage of reduction and solution of the deuteroplastic 

 contents being shown in fig, 6, while fig. 5 shows normal, 

 faintly granular cells in which the nucleolar body is absent. 

 Hecht (loc. cit.) notifies these cells as ' cellules speciales ', the 

 significance of which he discusses without offering a final 

 judgement. He draws them as loose cells and seems not to 

 have recognized their essential conjunctive nature, but com- 

 pares them with the large rounded or oval cells found in the 

 ceratal connective tissue of Galvina and other Aeolidiomorpha 

 previously described by Herdman (7). Comparison of sections 

 of animals at different stages of the alimentary cycle appears 

 to provide convincing evidence that both the secretum of the 

 cell-body and the refringent spherule of the nucleolus grow 

 during digestion and disappear during a fast. On account of 

 their structure, the readiness \yith which they take up both 

 basic and acid dyes, their position, on the one hand close to 

 the absorptive cells of the gut, and on the other on the walls 

 of the blood-spaces, and lastly on account of the significant 

 variation of their contents during a digestive cycle, it is here 

 proposed to regard them as protein storage cells. The agree- 

 ment in phase between the granular deuteroplasm and the 

 nucleolar secretion is in keeping with this explanation, and 



