408 TEGU.MENTARY ORGANS 



In Synoicum the test is soft, and prevents very much the structure of 

 some forms of rudimentary connective tissue. We find, in fact, a 

 more or less distinctly fibrillated basis with scattered endoplasts ; 

 some of these are invested by round granulous nitrogenous cell-walls, 

 while in others the cells are spindle-shaped and prolonged at each end 

 into fibres (representing thus the elastic element of ordinary connective 

 tissue), or they may be stellate. Botryllus, Syntethys, and Boltenia, 

 present a similar structure, varying, however, in the extent to which 

 the nitrogenous cell-walls on the one hand, and the periplast impreg- 

 nated with cellulose on the other, have undergone development. 

 Thus the periplast is broken up into very obvious fibres in Botryllus,. 

 while in Boltenia the fibrillation is pale and indistinct. On the other 

 hand, I have nowhere met with so great a development of the nitro- 

 genous cell-wall as in Synoicum. 



In Boltenia a more or less distinct lamination makes its appearance 

 in the test, and this peculiarity, as well as the fibrous structure alto- 

 gether, attains its maximum in the Cynthiae. In Cynthia papillata, 

 for instance, the middle substance of the test is composed of numerous, 

 very obvious lamina, which consist of fibres directed alternately 

 parallel with, and perpendicular to, the surface of the test (314. B.) 

 At first sight, they appear as Lowig and Kolliker have described them, 

 to be decussating sets of longitudinal and radiating ; but on a careful 

 examination of their sections I invariably found that the apparently 

 radiating fibres bend round as they approach the apparently longi- 

 tudinal set, and in fact pass into the latter. The longitudinal bands 

 are, however, no thicker at one end of a section than at the other, so 

 that the transverse fibrils cannot be merely given off from them. A 

 transverse section, again, exhibits the same appearances as a longi- 

 tudinal one ; so that I think the fibres must in reality have a more or 

 less regularly circular arrangement around the centre of the spaces, 

 occupied by the radiating bands, the apparently longitudinally fibrous 

 bands arising merely from the decussation of these circular fibres. 

 Great numbers of granular corpuscles (endoplasts ?) are scattered 

 through the midst of the " transversely fibrous " spaces. In Cynthia 

 pomaria, Lowig and Kolliker describe peculiar " cells " in the inner 

 layer of the test, consisting of such corpuscles surrounded by a thick 

 circularly fibrous wall, and the existence of these bodies appears to be 

 additional confirmation of the view I have taken as to the mode in 

 which the fibres in Cynthia papillata are disposed. If in the latter, 

 the fibres were disposed more closely around particular corpuscles, 

 the test would, in fact, break up into just such circularly fibrous cells. 

 I have hitherto described only the structure of the middle, most 



