19G 



HYDROID^. 



Part IV. 



Fig. 15. 



15, I; Fig. 12 and wood-cut 16, V') from one radiating canal to the other, so that 

 the four transverse channels, connecting the four radiating tubes, constitute, as a 



whole, the circular chy- 



niiferous canal of the 



medusa. The contract- 

 ing edge (wood-cut 13, 



€?) of the cujj, formed 



by the middle wall (a) 



in earlier phases, has, 



in the present stage, 



closed over, and forms 



a continuous wall (PI. 



XVIII. Fig. 12, and 



wood-cut 16, W'). In 



doing so, it has sepa- 

 rated the outer wall 



(wood-cut 16, (.(') from 

 its continuation, the innermost wall (c-*). These three walls {a^ V' c^) constitute 

 the transverse septum which shuts off the concavity of the disk from exterior 

 communication. The exterior wall (a) of the disk is still very thick, and the 

 innermost one (c) none the less so, but the middle Avail {h) is much thinner than 

 in earlier stages ; all three, however, are considerably thinner in the transverse 

 septum {u" b^ c*). The four sinuses (PI. XVIII. Fig. 10, and wood-cut 14, h% pointed 

 out in the last stage, have passed through the whole thickness of the disk, and 

 completely separated it into four lobes (PI. XVIII. Fig. 11, and wood-cut 15, o; 

 Fig. 12 and wood-cut 16, a)). Each one of these lobes, or young tentacles as they 

 may more propeily be called, is hollow to the very tip, and in direct communi- 

 cation with a radiating canal (wood-cuts 15, P, and IG, h^). These last are as yet 

 very broad and deep channels, whose walls occupy a large proportion, at least one 

 half of the thickness, and nearly the same amount of the circumference of the 

 disk. On account of the extensibility and contractility of the disk, this last pro- 

 portion is quite variable, as a glance at the two figures (PI. XVIII. Figs. 11 and 

 12) referred to above, will show. After this, the radiating tubes apparently diminish 

 very rapidly in diameter, and become gradually more slender (PI. XVIII. Fig. 13), 

 with the growth of the disk ; but in reality the}' increase not only in length, 

 but also in diameter, and the apparent reduction is owing to the more rapid 

 growth of the umbrella. The tentacles, also, correspond in rapidity of growth. 

 It will be noticed that they project centrifugally. so that their ends overlap 

 each other. 



