MUSEUM OF COMPARATIVE ZOOLOGY. 271 



resemble the nuclei of the epiblast as already described, are well 

 marked on the ridges of the scale. 



I have been unable to identify a scale of this kind with any of those 

 figured in the larval stages of Agalma as described by Metschnikoff. 

 Hccckel, however, figures a similar scale with divided tube in Crystallo- 

 des, but from his descriptions it does not follow that he regards it as the 

 modified primitive scale. In Fkysojjhora, however, we find an approxi- 

 mation in shape to this scale in the primitive hydrophyllium, and more- 

 over in this genus, as in mine, there is a smaller tube extending from 

 the cavity of the scale to the surface, and ending in or near clusters 

 of lasso-cells superficially placed. If the first-formed scales (primi- 

 tive hydrophyllia) in both Physophora and Agalma are homologous, we 

 may find the smaller bifurcations connecting the cavity of the scale in 

 Agalma with its surface to be the same as the similar structures de- 

 scribed by Hseckel in the young Physophora, provided, of course, that 

 the flat scale of Fig. 8 is the modified primitive covering-scale of PI. III. 

 fig. 14. The flat scale (fig. 8) is certainly different in the contour 

 and course of the central tube from the serrated hydrophyllia, and no 

 other structure is thought of to which to refer it except the primitive 

 hydrophyllium, that large covering-scale whose origin dates back into 

 the youngest stages of the larva. What has already been here written 

 of the modifications in form which the first-formed covering-scales go 

 through, does not of course show that in the end it may not be simply 

 cast off". My. studies throw no light on this point. If it is ultimately 

 dropped it undergoes modifications in outline before the consummation 

 of that event. 



Cambridge, July, 1885. 



