CERATA OF DENDRONOTUS. 363 



In fact, I am convinced of the correctness of the 

 description given in our previous paper of the minute 

 structure of the cerata, notwithstanding the doubt throv^n 

 upon it by Pelseneer. That description* is as follows : — 



"1. Large spaces in the mesoderm, containing blood 

 corpuscles {Log. cit., PL XII., figs. 2 and 3, c s). These 

 run in the main longitudinally. They occasionally branch, 

 and they open into innumerable minute lacunae in the 

 mesodermal tissues, all of which here and there contain 

 blood corpuscles. 



2. A good deal of pigmented connective tissue forming 

 branched masses and ramifying threads of a brownish 

 colour. These frequently, in a surface view of the 

 terminal branches of the cerata under a low power, give 

 rise to the appearance of a dark coloured granular central 

 caecum such as that figured by Bergh {Log. gH., PL II., 

 fig. 22). Sections, however, show the true nature of this 

 pigmented tract. 



3. Masses of large distinctly nucleated cells lying in 

 meshes of fibrous connective tissue. These are possibly 

 mucus secreting glands. They occur chiefly in the 

 smaller meshes of the cerata." 



I would respectfully suggest that a point of this 

 character, requires some of those " complicated methods of 

 histological investigation " which Pelseneer in his preface, 

 says he w-as unfortunately not in a position to make use 

 oL Without the aid of thin serial sections, I think it 

 would be impossible to determine with certainty this 

 point, and if, as was suggested in our previous paper, 

 these distinguished zoologists who differ from the con- 

 clusions come to, have worked without these aids to the 

 determination of minute structures, as I gather Pelseneer 

 has, then it is not surprising that they have been led 

 into error. 



Loc. cit., p. 230. 



