, PENNATUUDA. 



apex will always show that the upper wings are quite rudimentary in a more or less large number; not 

 only are they smaller than those farther down, although, if their polyps can be counted, they may show 

 the same number as these; but their polyps are quite degenerated, without tentacles, or with reduced 

 tentacles. The degree of degeneration increases upwards; where a somewhat large number of reduced 

 wings occur, the polyps of the lower ones may still be counted, then follow some in which only a 

 single polyp seems to be preserved, and finally, instead of wings we find but low and narrow trans- 

 verse ridges just as indistinct as those far down in the rudiment-zone. All these features I interpret 

 as indicating atrophy. As the atrophied part gradually dies the upper part of the calcareous axis 

 is quite naturally denuded; and this denuded part decays and is broken off, like the top of the 

 shell of certain snails with . truncate shells - - or is perhaps thrown off, after the lapse of some 

 time, as a stag throws off its dead antler. 



This process begins very early, as shown by the young stages described (in so small colonies, 

 that fishes cannot be suspected of having any share in the phenomenon), and it goes on -- normally 

 - hand in hand with the growth and increase of polyps in the rudiment region 1 ). 



This interpretation of the mode of growth in the Virgulariae is certainly not quite new, as I 

 thought at the time when I gave my first communication upon it 2 ); it has in reality been set forth, 

 contemporaneously with the first real description of Virg. mirabilis, by O. F. Miiller. In the Danish 

 edition of Zoologia Danica (1781) it is said of Straa-Sofjasren p. 45: When it has grown to a certain 

 height, the outermost polyp-hooks by and by die and wither away from above downwards at the outer 

 end of the stalk, and only the shrunken skin remains some time. Finally this also drops away, and 

 the end of the stalk is denuded. Accordingly, the polyps perish at one end as they are generated at 

 the other, which shows the second resemblance to the mode of growth of the tape-worm; yet the 

 stalk (i.e. the calcareous axis) remains entire and undamaged . Marshall has also had the same idea 

 (Oban Penn. p. 58), but he immediately rejects it, as he does not think that the facts sufficiently support 

 it. I think, however, such facts have been advanced here; and especially by pointing out the features 

 in the small young stages, I think I have placed this interpretation on a better footing at all events 

 than has hitherto been done. 



The young stages described are also of interest in another respect, as proving that two Penna- 

 tulid-genera (together with their family) will have to disappear from the system. One is the genus 

 Protocauhn of Kolliker (Rep. Challenger Penn. p. 26), with the species P. n/oltc, from a depth of 700 

 fathoms near New Zealand. My previous supposition, which was well-founded, that this species was 

 only a young Virgularia at a similar stage to that of V. mirabilis figured on PI. II, fig. 25, has been 

 fully corroborated by examination of the original specimen in the British Museum. The upper end 

 of the colony is not -- as might be thought from the figure I.e. PI. VII, fig. 23 -- provided with a 

 terminal polyp, but is imperfect after the manner of Virgularia; a small broken end of the cal- 

 careous axis projects distinctly. The polyps are solitary, in two alternating series as stated by Kolliker; 



•) If this explanation was not correct, and the condition of the top was a result of violence, it would be a 

 strange thing if specimens were never found showing that the colon}- has managed to repair the damage. Daly ell has 

 proved that Virg. mirabilis is possessed of no slight power of regeneration. On a living specimen D. cut off both ends, and 

 hung up the middle part on a thread; in three weeks both ends were repaired by regeneration (I.e. p. 1S6). 

 ) In the meeting of Naturhist. Forening on Nov. r i th 1S9S. See Vid. Medd. [898, p. 1. 



