CARADOCIAN CYSTIDEA FEOM GIRVAN. 503 



part of the stem stretching out at a low angle, and bearing at its end the theca, which 

 floated horizontally in the water, probably at no great distance from the bottom. 

 Under the influence of gentle currents, or in harmony with the efflux and influx of its 

 own anal stream, the theca swayed from side to side, hinged on the flexible region of 

 the stem, or from time to time rose and fell under the influence of a deeper wave than 

 ordinary. The outstretched arm, balanced by the antibrachial process, opened its 

 cover-plates and extruded its sensory podia, which, it may be, aided the ciliary 

 current of its groove. In this floating position, one face remained always uppermost, 

 and this face was a trifle more convex than the other, and was sometimes strengthened 

 by a faint folding of its plates. A slight amount of rigidity might also be given 

 by the thickening of some marginals on the outer edges of the lobes ; but this was 

 only just enough to stay the broad extensions and to prevent aimless flapping. 

 Thus the creature lived a peaceful life, leaving its anchorage only under the com- 

 pulsion of some storm that lifted it from the bottom or tore up the sea- weed in which 

 it was wedged. 



§ 586. If this picture be a true one, and I do not think it can be far out, then it 

 leads us to look at Rhipidocystis in a new light. This genus appears to be essentially 

 a Dendrocystid that has developed appendages at the sides of its stem. These, which 

 presumably have arisen in the same way as the cirri of more normally constituted 

 Pelmatozoa, are, in the middle region of the stem, like flattened pods, but near the 

 distal end are large, thin- walled sacks. . Dr Jaekel justly compares the latter to the 

 spheroidal, partitioned, hollow root of Scyphocrinus, but suggests, quite unnecessarily, 

 that the former may have borne the gonads. After seeing these curious structures, 

 I have no doubt but that Rhipidocystis, " like little wanton boys that swim on 

 bladders," floated in the water, supported by its swollen root-sacks ; not however 

 hanging vertically down from them, but kept in a more or less horizontal position 

 by the flattened paddles, as one might almost call them, of the main stem. The theca 

 swung more freely in the water than did that of Dendrocystis, and was therefore far 

 less flattened. 



§ 587. If the reader will kindly turn back to the account of the laterally keeled 

 columnals of Dendrocystis rossica (§ 1 54), he will find the details, previously so 

 unentertaining, now eloquent of the transition from the simpler stem of the Bohemian 

 species to the elaborate buoys of Rhipidocystis. Not that D. rossica itself was a link, 

 for it occurs at a higher level than at any rate the earlier species of Rhipidocystis [; but 

 the much older Dendrocystis from Herault also seems to have had side -vanes on some 

 of its columnals. June 1913]. 



§ 588. Having now attained a clear idea of the modifications in Dendrocystis, let 

 us try to ascertain what sort of a creature it was that became modified. In considering 

 the brachial appendage (§ 72), I have given reasons for supposing that this did not 

 exist in the normal pelmatozoic ancestor. That ancestor would naturally have had at 

 least three, probably five, and possibly more brachioles, or else none at all. If it had 



TRANS. ROY. SOC. EDIN., VOL. XLIX. PART II. (NO. 6). 66 



