154 TEANSACTIONS LIVERPOOL BIOLOGICAL SOCIETY. 



Now the points insisted on as distinguishing Bepastrnm 

 from Carduella were these : — that while in both the ten- 

 tacles are arranged in 8 groups round the margin of the 

 disc oral surface of bell or umbrella) yet in Depas- 

 ti'iim the tentacles are very numerous, and are arranged 

 in each group in several rows or series, one within another, 

 and further that they spring from the margin of the 

 octagonal disc or from without it, but that in Carduella 

 there are only about 5 tentacles in each group arranged 

 in a single series and arising completely ivithin the margin 

 of the circular disc. Further in Carduella there is a single 

 tentacle (primary tentacle of Haeckel, corresponding to the 

 marginal anchors of Haliclystus) in each interval between 

 the groups of other tentacles ; this does not seem to have 

 been observed by Gosse in Depastrum. Clark pointed out 

 that, according to the figures and description of Sars, con- 

 firmed and extended by his own examination of the 

 specimens sent to him by the latter, the Norwegian species 

 when adult has the tentacles arranged in several rows 

 (3 or 4), but that the younger individuals have at first 

 only one row and later two. Further he pointed out that 

 Allman was in error in asserting that the Orkney form 

 always had its tentacles in one row only, for some of those 

 sent to him, being presumably older than those described 

 by Allman, had more than one row. He accordingly con- 

 cluded that Carduella as defined by Allman w^as merely 

 the young form of the Lucernaria cijathiformis of Sars. 

 With regard to this point, the examination of a large 

 series of specimens at Port Erin has amply confirmed the 

 conclusions of Clark. Now as to the Depastrum of 

 Gosse, Clark inserts it provisionally as a distinct genus, 

 but at the same time he gives some very good reasons 

 for regarding it also as identical with Lucernaria cyathi- 

 formisj Sars. The difference in the number and arrange- 



