676 MK. A. W. WATEKS ON MEMBKANIPOKID^. 



of the spines inside the zooecial tubes of LicJienojpora *, Entalo- 

 phora, and other genera of Cyclostomata, and there are also 

 spines projecting from the internal walls of Flustra dentigera, 

 denticulata, and spinuUgera. 



There are four or six lateral rosette-plates, and nuiuerous 

 distal ones at the base of the distal wall. The appearance differs 

 considerably in various parts of a colony and in different colonies, 

 as the tubercles are sometimes small, at other times very large, 

 and two frequently merge into one, as figured by d'Orbiguy. 



Bijiustra jugalis, MacGillivray, seems to be closely allied, 

 and if M. Hyadesi, Jullien, has to be separated, it would only be 

 as a variety. M. danica, Levinsen, belongs to this group, also 

 M. Umamillata, MacG., a species which I have not seen, and I 

 do not understand what MacGrillivray says about the denticle. 

 M. denticulata, Smitt (Floridan Bryozoa), may be a variety. 



There is a tendency for the lateral walls to be broken through 

 by the action of the waves, but they frequently become cemented 

 tot^ether again. The same is the case in M. memhranacea (PI. 49. 

 fif. 17), and in both species the break occurs in approximately 

 the same position. In neither do I see any trace of the separate 

 plates which Nitsche describes. 



Log. California (given by Miss Jelly) ; Angola ; Chagos 

 Island; San Pedro, Pernambuco ; Senegal; Kurrachee {Brit. 

 Mus.); Eio de Janeiro (MacG.); Madeira (Busk); Patagonia 

 (d'Orb.). Gulf -weed generally. 



Membeanipora tehuelcha, var. inteetubeeculata, nov. 

 (PI. 48. figs. 1, 2.) 



Specimens from Port Elizabeth, S. Africa, have the border 

 very much raised, so that the membrane carrying the operculum 

 is much depressed. On the lateral wall, besides the tubercle 

 there is a projecting process, and it is upon this and the de- 

 pression that the variety is based. 



The tubercles vary much in shape and character, and it is 

 almost impossible to give a faithful representation. The walls 

 are much thinner than in the figures, which must be looked 

 upon as somewhat diagrammatic to show the main characters. 



The minute denticles or rays occur near the base as in typical 

 tehuelcha. There are no ovicells or avicularia, and on the lateral 

 wall there are four rosette-plates. 



* See Waters, " Tert. Cycl. Bry. from New Zealand," Quart. Journ. Geol 

 Soc. vol. xliii. p. 340, pi. xviii. figs. 5 & 6. 



