Ig LUCERNAKI^ AND THEIR ALLIES. 



(fig 3) or they are all alike. The first case, though, is the one which is most 

 generally met with, and 'sometimes the shorter sides are so strongly contracted and 

 narrowed that the umbel has rather the appearance of a quadrilateral with four 

 double corners {fig. 22), or the corners project so far as to have rather the appear- 

 ance of arms (AV- l"^)- '^^^ese sides are always more or less inarched, and with 

 such a uniform" curvature that tlie middle of each broad sinus is the point nearest 

 to the proboscis, and consequently tlie borders of the same are most distant, and 

 combine to form tlie corners of the octolateral. 



42. Umbellar Appendages.— A. very strong emphasis is put upou these alternating 

 corners and sinuses by the appendages which project from their edges. The corners 

 are rendered apparently more prominent by the implanting thereon of a thick group 

 of cylindrical, globe-tipped tentacles {figs. 1-17, 22, 37, (^), whilst the comparatively 

 inconspicuous edge of each interval is intensified by the addition of a dark, oval, 

 kidney-shaped adherent organ (a). The manner of attachment of these bodies will 

 be described under their respective headings, as we do not wisli here to complicate 

 tlie subject any more tlian is necessary to render the relation of the various parts 

 of the umbel sufficiently clear for a perfect understanding of topography and mor- 

 phology. The adherent organs (anchors) from their position in the sinuses, form 

 a marked feature in the subdivision of the octagon, inasmuch as they lie severally, 

 either opposite the flat sides of the proboscis or exactly confronting its four corners, 

 and, therefore, have a closer and more direct relation to the planes of bilaterality 

 and dorso-ventrality than the tentacles possess, which, as it were, stand obliquely to 

 these, at regularly alternating points. 



43. The hngUudinal diameter of the umhdla, i. e., the distance from the anterior 

 to the posterior face, is quite diverse at dift'erent points, but not without system. 

 We have already anticipated this by inference, in the description of the basal 

 prolongations of the proboscis (39). Conjoined with the four proboscidal but- 

 tresses (^), the four equidistant regions which lie opposite the angles of the 

 proboscis have a far greater antcro-posterior extension from the front to the back 

 face than any otiier portion of the umbel, whilst the middle of the four subdi- 

 visions, which abut against the flat sides of the proboscis, measures the least 

 in this respect. The reason for this will be apparent enough without going into 

 details, upon stating that the middle of each of the four sunken areas (39) 

 corresponds to tlie line along which the walls of the anterior and posterior foces 

 of the umbel are united. These lines of junction— internal partitions (4--). as 

 will be learned hereafter (47) -may be recognized from without as four narrow, 

 light bands trending, severally, from each of the four sides of the proboscis almost 

 to the margin of the umbel' Consequent upon this it is plain that the antero- 

 posterior faces are further apart at one intervening point than at the partitions, 



' One of the most elegant and characteristic figures of this species thus far published (sec A. 

 Agassiz's Illustrated Catalogue of the Museum of Comparative Zoology, No. II., North American 

 AcalephjD, Cambridge, Mass., 1805, p. 03, fig. 88) is unfortunately marred by a serious morpho- 

 logical mistake in the drawing, by which the angles of the proboscis are made to appear as if lying 

 oppos.te_ the partitions of the umbclla and consequently facing toward the four double reproductive 

 organs, uistead of toward four points intermediate between these. 



