Chap. I. REPRODUCTION OF CORYNE MIRABILIS. 



191 



(PL XX. Fuj. 4, t) appear, nearly oppcsite to each other, on the head, and not far 

 below its apex. These knobs are densely crowded with lasso-cells, which give 

 them the appearance of being the globular tips of the tentacles. A bud a little 

 older (PI. XX. Fig. 5) not only discloses the nature of these knobs, but also shows 

 that they belong to a hydroid form exactly like the stock from which it arises. 

 The knobs of the last phase have become elongated on a short i)edicel, and broad- 

 ened a little, and two more of the same kind have grown out nearly opposite 

 to each other (PI. XX. Fig. 5, t\ at two points ninety degrees from the situation 

 of the first two, but a little higher up on the head. They have now every 

 characteristic of tentacles, and unmistakably demonstrate that the globular tip of 

 the tentacle is developed first, as suggested in regard to the earliest phase (PI. 

 XX. Fig. 4, i). At this stage the young Hydroid resembles the genus Stauridia 

 of Dujardin,^ in the relation of the tentacles to each other, and to the head, on 

 which they are so arranged as to resemble a Maltese cross, when seen from above. 

 We have already alluded to the probability that the tentacles do not develop 

 absolutely by twos and multiples of two, but so closely one after another, and 

 so nearly on the same level in the early stages, that they have the appearance 

 of originating in pairs. In a not much further advanced phase, there are six 

 tentacles (PI. XVII. Fig. 8), four of them arranged as in the last stage, and two 

 higher up, which appear to be jjlaced at intermediate points, over two opposite 

 angles of the lower cross. That the tentacles are very irregular, at times, in their 

 development, may be seen in a figure of a young Hydroid on which eight tenta- 

 cles (PI. XX. Fig. 6, t) were counted, all of them mere knobs, filled with lasso-cells. 

 We have no doubt that in this case the tentacles were all pretty nearly equal 

 in development, and, moreover, just beginning to bud. By the time the Hydroid 

 has ten tentacles it may be considered as adult, if we may judge from the fact 

 that it may bear medusae (PI. XVII. Fig. 9, md). In the oldest Hydroids which 

 we have seen, the tentacles very seldom exceed sixteen in number (PL XVII. 

 Figs. 11, \Y, and 12). 



Since the Hydroid never buds from any other part of the parent except the 

 stem, it must of necessity pass through the horny tube, in order to be able to 

 develop. It does not, however, make an open passage through the tube (PL XX. 

 Fig. 3, c), but absorbs the horny substance where it touches it, and at the same 

 time elaborates a thin sheath (PL XX. Fig. 3, c^) for itself, which is united to the 

 edge of the opening in the old tube, resulting in a continuity of the two. Before 

 the formation of the mouth, in the young Hydroid, the new sheath remains a 



» Annales des Sciences Naturelles, 1843, Vol. Figs. Q} to C. Cladonema Dujard. (later Stau- 

 XX. p. 370, and 1845, Vol. IV. p. 271, PI. XIV. ridia Wright) is the free Medusa of Stauridia, 



