110 



LUCERNAKI^ AND THEJR ALLIES. 



Tlic i)osition of the fibrilliE, however, needs special mention. They hold the 

 same o-eiicral relation to the adjoining layers as in the unibella, but in the latter 

 they all point or trend one way, i. e., antcro-posteriorly, while in the tentacles they 

 vary, not in reference to the ectophragma and endophragma (gastrophragnia), but 

 following the curvatures of tlie latter as they form the hollow shaft, they always 

 trend in the direction of the radii of these curves. As a consequence of this, then, 

 their trends radiate from the axis of the shaft, except at the tip of the organ, 

 where the layers converge and form a series of superposed hollow hemispheres, 

 and there the trends of the fibrillae radiate as if from the centre of a sphere. Their 

 trend in the iatertentacnJar lobules will be best comprehended in connection with 

 the descrii)tion of the morphology of these bodies, to be found in a previous para- 

 graph (^I 103). 



205 (A). Over a greater part of the regions in which the chondromyoplax pre- 

 vails its fibrilla; abut at right angles against tlio adjoining layers, but at the bases 

 of the tentacles their disposition is modified in a marked degree, in connection 

 with the peculiar arrangements of the muscular layer. The latter, as described in 

 a former paragraph (^[ 82, 100) with considerable detail, passes obliquely from the 

 periphery to the inner layer, or gastrophragnia, forming a partition {fi(j. 60, A') 

 between the chondromyoplax (6') and the edge of the chondrophys (e). The 

 fibrilla:' of both tliese layers abut against this oblique partition, at very acute angles, 

 tlieir proximal ends bcins; based on the gastropliragma («, r). This we should 

 say, also, takes place along the wliole circuit of the edge of the nmbella, wherever 

 the chondrophys and chondromyoplax adjoin a common area of termination. 



206. Tlte Gafitrophmgnia {endojjhragma) (J^] 101) of the tentacles (Jigs. 90, 91, 

 f ). Tliis single stratum of cells owes its great thickness only to the extreme 

 depth of tlie latter. It varies considerably according as the tentacle is extended 

 or contracted ; in the former case the cells, lessening to their minimum depth, are 

 still three or four times deeper than broad, and in the latter case they have a 

 much greater measurement. They do not, therefore, exhibit that wide range of 

 variation in proportion which the cells of the ectophragma (n^) of tlie shaft do, but 

 have rather tlie approximate fixedness of tliose of the globular tip; and, as they 

 are not so irregular in outline at the sides, they are more prismatic in appearance. 

 In a profile view (///• 9^') of a tentacle these cells present parallel sides, but in a 

 cross sectional view {Jhj. 91, r) of this organ, they very naturally, from being 

 arranged radiatingly about an imaginary axis, appear narrowest at their proximal 

 ends, and gradually expand, Avedge-like, to their distal terminations. They are 

 not, though, four-sided, as these two views might, perhaps, lead us to suppose, 

 but irregularly polyhedral, varying from tliree to six sides, as may be learned by 

 looking at them endwise {pj. 90, 1 r). Their walls are about as thick as those 

 of the ectophragma {n'), and, like those, very sharply defined within, but 

 rather obscurely without, either because they are overlaid by the intervening pig- 

 ment granules (e=) about their proximal halves, or from being more or less organi- 

 cally united along their lines of contact, at their distal halves. The rigidity of a 

 prismatic conformation is negatived by a long undulating contour, which rather 

 enhances, than lessens, the brilliancy, by the ever-varying surface which tlie walls 



