058 ME. A. W. WATERS ON MEMBRANIPORID^. 



A number of species have vicarious avicularia, wliicli Busk* 

 defines as avicularia which " represent or replace an ordinary 

 '/ooecium," in contrast to adventitious avicularia, which are 

 •' either attached to some part or other of a zooecium or inter- 

 spersed among the zooecia." 



Although it is useful to speak of vicarious avicularia, it is 

 impossible to draw a hard-and-fast line. In such forms as 

 M. crassimarginafa, Hincks, M. perfragilis^ Hincks, the avieu- 

 larian chamber is the same shape as the ordinary zooecia, and may 

 well be called a zooecial avicularium, whereas in M. pyrula, H., 

 and 31. permunita, H., the avicularian chamber is much smaller 

 and of a different shape. Erom these to the avicularia of 

 M. tenuirostris^ &c. is only a step, and it would seem best to 

 consider vicarious avicularia those in which the avicularian 

 chamber extends to the basal wall, dividiog them into vicarious 

 similis, and vicarious dissimilis. Adventitious avicularia are 

 those which are placed upon the zooecium, as in M. spinifera, 

 Jobnst., or in M. ecTiinata, d'Orb. 



There are vicarious avicularia in the genus Onycliocella, and 

 also in M. crassimarginata, M. pyriila, M. valdemunita, M. per- 

 fragilis, M. longicornis, M. papuUfera, M. onarginella, M. nolilis, 

 M. aperta, M. corhula, M. dissimilis, M. sceletos, M. velafa, 

 M. jlustroides, M. maderensis. 



The avicularia are articulate in M. Garteri, and pediculate 

 ia M. cristata, M. cy mice for mis, M. spinifera, M. ecMnata, and 

 M. echinus. In M. elUptica and M. imlricata the avicularian 

 opening is very narrow. 



The rosette-plates have been examined in a number of species, 

 but of many I have not had suitable material, and where par- 

 ticulars are given from the examination of small fragments, 

 which could not be broken up, it is possible that a mistake 

 may have been made even although reasonable care has been 

 taken. 



In some Memlraniporce the rosette-plates are simple disks in 

 the zooecial walls as in Flustra ; in others, as already mentioned 

 by Levinsent, there are chambers at the base of the zooecium 

 into which the protoplasmic threads pass from the neighbouring 

 zooecia. The pore-chambers {porelemnre) must be looked upon 

 as equivalents of the rosette-plates as far as their number and 



* Zool. ' Ohall.' Exped. pt. xxx. p. xviii. 

 i' Danske Dyr (Mosdyr), Zool. Danica, p. 7. 



