14 



another, down the " basilar column " ; there is no common region except the 

 base ; the large polyps expand from the top of the basilar column, but there is 

 no branched or lobed poly])arium ; in fact, as the author says, there is no 

 polyparium properly so-called. He makes the same remark, it is true, in regard 

 to Pai-alcyoniiim, which, however, he had not seen. What at once marks Paral- 

 cyonium as distant from Fagcwularia, is the presence of a branched polyparium 

 rising from the top of a firmer cylindrical stalk, into which it can be retracted. 

 Viguier speaks of the " incontestable resemblance " between his Fasckulana 

 and the Parahyoni/im of Milne-Edwards, but we are unable to share this view. 

 The description of Fuscirtilarm suggests to us relationship with Sympodium 

 rather than with Paralcyonimu. 



A recent careful study of abundant material of Fasciculuria and Paralcy- 

 onium by Sophie Motz-Kossowska and Louis Fage ("Contribution a I't^tude de 

 la famille des Fascicularides," Arch. Zool. Exper., vii. (1907), pp. 423-43 (10 

 figs.)) corroborates Viguier's view. In their interesting pajjer the authors point 

 out that the two types agree (1) in having a stolon connecting the colonies (but 

 this is often almost suppressed in Pandcyouinm) ; (2) in having a rigid basal 

 portion into which the polyps can be retracted (but in Fascicularia this is com- 

 posed of the unfused gastric cavities of the polyps, wiiereas in Paralcyoniiim 

 there has been much coalescence, and therefore far fewer longitudinal canals 

 than polyps ; moreover, Paralcyon'mm has a branching polyparium with second- 

 ary polyps arising from primary polyps) ; (3) in having similar spicules — small 

 flat, opaque elliptical forms in a sub-tentacular collar and larger spindles in the 

 basal portion (but the spindles are very much larger in Paralcyoniiim). The 

 authors point out that I'dndr.yonimn passes through a Fascieidaria stage, and 

 in spite of the great difference in the basilar portion and in the relations of the 

 polyps to one another, they unite them in the family Fascicularidtt, defined as 

 follows : " C'olonies very poor in ccenenchyma, composed of several groups of 

 polyps united by a stolon ; polyps united at the base in a rigid column within 

 which they can be completely retracted ". It is suggested that the family is 

 connected by Fascieularia with the Clavularidai, that there are some affinities 

 with Nidalia and Nidaliopsis, and that the nearest related form is Onjaiiklm. 

 The armature of the polyps in Nidalia, its unbranched habit, and many other 

 features separate it far from Paralcyonlam, and Ger^emia, to which, as Kiiken- 

 thal has shown, Oiya/iidax must be referred, is equally remote. 



We have not as yet been able to procure a specimen of Fascieidaria for 

 examination, and we would not therefore dogmatically exclude the possibility 

 that Fascieularia, Paralcyoniuia and Stiideriotes form a series showing the jiro- 

 gressive differentiation of a rigid basilar portion into which the rest of the 

 colony can be retracted. The descriptions given by Motz-Kossowska and Fage 

 are very precise, and their discussion of the possible relationships is admirable ; 



