Chap. IV. PARYPHA CROCEA. 261 



they are more or less rounded, so that the surface of the wall appears papillate. 

 If a piece of a tentacle is cut off, it soon disintegrates, and these cells assume 

 a spherical shape {Fig. 1% ABC). In this condition their contents may be studied 

 with great facility. When thus isolated, one can determine absolutely that the 

 lasso-cells {a b P c d) are within the cavity, or at least within the superficies of 

 the cell. There is good reason to believe that the lasso-cells are not strictly 

 within the cavity of the cell, but are imbedded in the thickness of the wall ; but 

 this is so extremely transparent that we have not been able to define its inner 

 boundary with certainty. We have succeeded much better with the cells of the 

 outer wall of the stem {F/(/. 4, b), where the cell wall {Fiff. G, c) appeared to be 

 quite thick, and, at the outer end, so incrassated (at b) as to include the whole 

 length of the lasso-cell (a) Avhich stood transverse to the surface. This thickening 

 may not be, properly, a part of the cell wall, but a lining of it; on this point, 

 however, we are much in doubt. 



A larger part of the contents of the cells of the outer wall of the tentacles 

 are coarse, irregular, granular, oily-looking bodies, some of which {Fi//. 1", e) are as 

 large as the lasso-cells. The lasso-cells belong to the same type as those found in 

 Hydractinia polyclina (PI. XVI. Figs. 10 and 11), to the description of which we 

 will refer for the details in regard to these bodies. The inner, or axial layer of 

 the lower tentacles, is composed of very peculiar cells {Fiffs. 1, c, 2, e, and 3, e). 

 Upon taking a profile view of the tentacle, these cells, at first glance, appear to 

 be irregularly and sharply polygonal {Fiff. 1, e), but, by plunging toward the centre 

 of the tentacle, we find that they have a much greater diameter transverse to 

 the axis than along the same, and hence that they are four or five-sided jjris-- 

 matic bodies. They do not all converge toward a central line, but trend parallel 

 with each other, and extend from a plane which is parallel to the flat side of 

 the tentacle, to the plane of the axis. In a view from above or below {Fiff. 2, e) 

 the tentacle, their longer diameter is displayed, showing, in a more direct way, 

 their elongate prismatic form, and also another peculiarity, not easily to be discov- 

 ered from any other point, namely, that they all have a greater or less inclination 

 toward the tip of the tentacle, so that the two roAvs of superimposed cells meet 

 at an obtuse angle, in the perpendicular plane (e^) of the axis. By taking 

 advantage of the bending of a tentacle in its multiplied contortions, one may get 

 a view of a most perfect transverse section {Fiff. 3). The opposite sides of these 

 cells are not often parallel with each other, one side making two or three curves 

 in its course, while the other side makes but one or two curves, or sometimes is 

 nearly straight; and again, the ends alternate irregularly, one end being broader or 

 narrower than the opposite one. The walls are much thicker than those of the 

 outer layer of cells, but we have never been able to see each one singly, so 



