--. DISCOVERY REPORTS 



254 



text-fig. i) in ^. guntheri. In addition to the autozooids forming the whorl, three fuUgrown autozooids 

 are found on the inner side of the ring, evidently displaced or pushed to the inner side of the whorl 

 at intervals between autozooids of the outer circle. This dis- 

 placement may possibly be due in part to a contraction of the 

 rachis, especially in the case of the single displaced autozooid 

 (unstippled in text-fig. 2). But the two inner autozooids on the 

 other side are more likely to have budded inside the whorl. As 

 far as we can judge from the literature, this specimen is the 



largest U. thomsonii examined up till now. This may explain ■ — y.^ /"^ 



the ' irregular ' arrangement of the autozooids as compared with f J t^'^\) 



Broch's (1957) specimens. Probably very large clusters will be \^ 



found to exhibit a more or less complete inner whorl of auto- \ ,-' 



I I 



ZOOlds. Pjg 2. UmbelMa thomsonii Koll. (Specimen 



The re-examination of Kolliker's type thus confirms the a, lectotype). Diagram of the arrangement 



synonymy of U. thomsonii and U. giintheri, and the first name of the autozooids in ventral aspect. The 



must therefore have the priority. primary polyp hatched, the two autozooids 



(Lectotype: British Museum (Nat. Hist.) Reg. No. 1881. which have been cut away st.ppled. 



2. II. 23. Kolliker's Specimen A designated on p. 253, paragraph 4 as the lectotype). 



NON-SPICULATE SPECIES OF Umbellula 



In this second group of Umbellula, variations in, and contraction of specimens make the limitation of 

 species exceedingly difficult, and specialists have wavered in their interpretations between a couple of 

 species only, or a vast number, describing practically every new specimen as a new species. Here, of 

 course, degrees of contraction have been welcomed as ' specific criteria '. 



In these non-spiculate forms also, the shape of the axis is a comparatively reliable character, but 

 again it is necessary to study the naked axis and not to judge by a superficial survey when the 

 axis is enveloped in its more or less fleshy, inflated or contracted coverings. 



Apparently nobody has studied the axis of the Umbellula species with the aim of determining the 

 role of the calcareous incrustation. In taxonomic literature, among the specific characters often given 

 is the possession of either a 'rigid', or a very flexible, stem. (It should be mentioned here that living 

 pennatularians and their axes are throughout much more flexible than preserved specimens.) Where 

 the calcareous incrustation of the axis is very scanty or almost wanting, the horny structure in some 

 cases is rather more spongy than flbrous, and in such cases compression may distort it to a certain 

 degree. The anatomical structure of the surrounding sheaths of the holdfast (or stalk) and their system 

 of canals may then give the false impression that the axis was approximately quadrangular, and it is 

 essential to make a critical study of the axis itself, its structure and the degree of calcareous in- 

 crustation. 



The non-spiculate species fall into two groups : A, with a round axis, only very little if at all incrusted 

 with lime, and B, with a quadrangular axis rather heavily incrusted with calcareous matter. 



Group A 

 Axis round, devoid of lime, very flexible. 



Umbellula huxleyi Kolliker 1880 (PL i. Figs, la, b) 



U. huxleyi pars Kolliker 1880, p. 21, PI. 9, figs. 37 a, b. 



U. gracilis A. M. Marshall 1883, p. 142, PI. 25, figs. 29-35. 



U. huxleyi Nutting 1909, p. 711. 



