446 Prof. W. J. Sollas on the 



small ones is to be seen, as if resulting from the fission of one 

 of the unusually large forms (PI. XVII. fig. 21). 



Mesoderm. — This consistsof a gelatinous connective tissue, of 

 which the matrix is a quite colourless transparent jelly, highly 

 unalterable by acids and alkalies, and remarkably poor in gra- 

 nules, those present being exceedingly minute ; its corpuscles 

 (PI. XVII. figs. 25, 29, 30) consist of a variable quantity of 

 granular propoplasm, often vacuolated, and provided with an 

 oval or round nucleus 0"00013 to 0"00017 inch in diameter, 

 within which is a minute nucleolus. The outer protoplasm 

 extends into long branching processes, which terminate in 

 threads, scarcely traceable near their ends for fineness. Some- 

 times the threads diminish regularly up to their ends ; some- 

 times after diminishing they thicken out up to a point of 

 bifurcation (fig. 30) ; frequently the angle of the bifurcation 

 is filled up by an accumulation of protoplasm ; some- 

 times, finally, a short process from the corpuscle thickens 

 into a lump of sarcode at the end, from which several short 

 hair-like processes radiate outwards (PI. XVII. fig. 30). 

 Sometimes the fine ends of the threads appear to terminate 

 freely ; more often they unite with those from neighbouring 

 corpuscles. A large proportion of them are elongated in one 

 direction and joined end to end to form long granular nucle- 

 ated threads (PI. XVII. fig. 25) ; the lateral branches pro- 

 ceeding from the protoplasm about the nuclei of the corpuscles 

 unite with similar threads or enter other corpuscles. Some- 

 times the matrix about the fibre becomes in places finely 

 fibrillated parallel with it (fig. 25,/). The ends of the fibres 

 or of the branches from them appear to be ultimately brought 

 into close connexion with the ectodermic and endodermic 

 layers ; for on the inner faces of these layers fine filamentous 

 processes are often seen wandering, and the branching filaments 

 of connective-tissue corpuscles can frequently be traced right 

 up to them ; in several cases also, I believe, I have seen a 

 connexion between the individual cells of a flagellated cham- 

 ber and the branching processes of a corpuscle (fig. 15). It 

 is, indeed, difficult while studying this reticulum of connec- 

 tive-tissue corpuscles to resist the idea that we are here deal- 

 ing with something that plays the part of a nervous system. 

 And just as the nervous tracts usually follow and are pro- 

 tected by the skeletal structures, so here a large number of 

 the corpuscular fibres are seen running parallel close by the 

 side of the chief spicules of the body. On the other hand, 

 the modifications which some of the corpuscles undergo seem 

 inconsistent with special nervous properties. 



In an irregularly defined layer a little below the skin, at 



