82 THE OCEANIC HYDROZOA. 



The endoderm and ectoderm were exceedingly distinct in the walls of the young 

 hydrocysts. There were no villi ; but the apices of the organs contained several large 

 thread-cells. 



Vogt does not mention or figure any small tentacular appendages to the hydrocysts, 

 and KoUiker distinctly states (p. 22) that he " could find no trace of the fine tentacles 

 observed by Philippi attached to these organs in P. tetrasticJia." In the young forms 

 of my specimens, on the other hand, they were very obvious as delicate, filiform pro- 

 cesses of the ectoderm and endoderm of the hydrocyst, which arose from the upper wall, 

 close to its attachment to the coenosarc. The internal cavity of each tentacle communicated 

 freely with that of the hydrocyst, and towards its apex delicate thread-cells were imbedded in 

 its outer wall. I did not particularly examine the polypites, but Vogt and KiJlliker agree in 

 stating that they are affixed by slender pedicles, and that the tentacles are attached at 

 the junction of these pedicles with the enlarged digestive division of the organ. Villi are 

 developed from the endoderm of the latter. 



The stem of the tentacle^ is as usual single, and carries a number of lateral branches, to 

 whose extremities very singular saccular bodies are attached, of whose structure I can speak 

 with some confidence, as I worked out the principal steps of their development. 



In their early stages these lateral branches exactly resemble those of the tentacles 

 of the Cali/cophoridce, that is to say, they commence as simple, csecal, double-walled 

 proces,ses of the stem, which gradually elongate, and become divided into three portions. 

 The one wall of the middle division becomes greatly thickened, and rows of elongated thread- 

 cells disposed vertically to its plane are developed in it. The distal division remains short, 

 and acquires only small thread-cells. 



The proximal division elongates and becomes a slender pedicle, devoid of thread-cells 

 in the greater part of its length. When it joins the middle division, however, it becomes 

 dilated, and not only small thread-cells, but several very large oval ones, appear in it. 

 Similar thread-cells make their appearance in the proximal end of the middle division. Thus 

 far the development of the organ (which has now a length of about one fortieth of an inch) 

 exactly resembles that of the corresponding part in Diphi/es, and it is obvious that the basal 

 division answers to the pedicle and involucrum, the middle division to the sacculus, and 

 the distal to the filament (fig. 5). 



The next change that takes place is, that the sacculus lengthens and becomes coiled upon 

 itself, while the involucrum at the same time dilates and acquires a quadrate outline (fig. 6). 



In the most perfect organ (which measured about one sixtieth of an inch in diameter, 

 but was still colourless, and, consequently, incomplete) the involucrum has become 

 immensely dilated, so as to form a sort of hemispherical cup closed on all sides, except 

 at one point, where a small aperture is left, through which the extremity of the filament, 

 divided into three lobes, protrudes" (figs. 7, 8). 



* Attaining as much as a foot in length in P. tetrasticha, Philippi, 1. c, p. 61. 



^ Compare Philippi's fig. 9. The lobes are described iis "a few short cirri." But this observer 

 has not comprehended the structure of those singular organs, and while he definitely asserts they are 

 not prehensile organs, supposes tliey may be branchise. 



