Mr Men on, Notes on Semper' s Larvae. 413 



numbers in the outer and middle parts of the ectoderm and are not 

 confined to the deeper layer. Below the nucleated layer of the ecto- 

 derm there is a zone of granular material which is closely applied 

 to the mesogloea and which remains colourless in stained prepara- 

 tions. From the layer immediately outside the mesogloea which 

 contains most of the nuclei belonging to the granular zone, con- 

 spicuous strands of granular material, with nuclei here and there, 

 proceed inwards through canals in the mesogloea. They can, 

 often, be traced right across the mesogloea into small granular 

 masses lodged in depressions on its endodermal side. Cords of 

 the same material are also seen ascending into the mesogloea of 

 the mesenteries. 



The mesogloea is a well-developed homogeneous structure. It 

 is thicker in some places than in others. Its inner border is very 

 irregular, being often raised into broad conical processes project- 

 ing into the endoderm between the mesenteries. It does not 

 show any lamination, even where it is thickest. It is pierced by a 

 network of spaces in which the above-mentioned strands of 

 granular material are lodged. Small granular masses are seen in 

 sections of the mesogloea and are no doubt sections of these 

 strands. In some specimens — of presumably a later stage — these 

 cords are hollowed out to form canals. These canals, with 

 bridges of granular substance running across them, are seen in the 

 mesogloea of the mesenteries, both large and small, as well as in 

 the parietal mesogloea opposite the attachments of the mesen- 

 teries. In the earlier stages of the larva the mesogloea is ex- 

 tremely thin and does not contain these elements. But their 

 connections in the later stages render it almost certain that they 

 are derived chiefly from the ectoderm. Similar ectodermal strands 

 and canals have also been described in Zoanthus by Hertwig. 



The longitudinal muscles of the mesenteries are much better 

 developed than in Zoanthella. Their arrangement is the same as 

 that already described for that form. 



The endoderm is for the most part a vacuolated layer, con- 

 taining numerous scattered nuclei and yolk masses of different 

 shapes and sizes. Round the stomodaeum it forms a thin layer; 

 its parietal portion is much thicker and increases in thickness 

 aborally. The yolk masses which are abundant in some specimens 

 are comparatively few or absent in others. Curiously enough the 

 youngest larva had hardly any yolk in it. In the peripheral layer 

 of the endoderm there are often seen small or large coarsely 

 granular yellow particles. They are conspicuous in specimens 

 with little yolk and are probably the remains of the broken-down 

 and absorbed yolk spheres. The endoderm is not so highly vacuo- 

 lated and does not fill up the coelenteron or encroach on it to the 

 same extent as in Zoanthella, 



