84 Mr. A. W. Waters on Australian Bryozoa. 



valuable for diagnostic purposes, I propose to designate it the 

 " lucida." 



There is another character which seems to have had no 

 attention ; that is the method of rooting or attacliment. In 

 some, as C. cacatua^ G. delicatula, 0. alata, G. carinata, G. 

 ^usilla, G. jmlchella, an isolated chitinous tube starts from 

 the back of some of the globuli, and is attached by a spreading 

 grapnel to any substance near ; in others, as G. crystallina, 

 C. formosa^ G. crihraria^ G. taurina^ G. cornuta^ G. perfo- 

 rata, G. Hannafordi, G. elegans, G. insignis, G. ventricosa, 

 G. hastata ?, G. lorica, tubes arise on either the dorsal or front 

 surface, sometimes on both, and several such radicles may 

 start from the same zooecium. These unite and form solid 

 bundles, becoming thicker towards the base. So far as my 

 collection enables me to judge these are from the dorsal sur- 

 face in G. Hannafordi, G. margaritacea, G. ventricosa, G. 

 hastata ?, G. lorica, G. formosa, G. crihraria, G. cornuta, G. 

 perforata, and from the front in G. elegans ; in G. umhonata 

 mostly from the front, but also from the dorsal surface ; in G. 

 deUcatula, G. insignis, G. taurina from the back or front. 

 How far these growths are influenced by local conditions can 

 only be examined by those on the spot. 



Rooting and articulation seem to be correlated growths — 

 that is, chitinous tubes may be given off to attach the colony 

 to foreign substances or to attach one internode or one part 

 to another, thus allowing motion without destroying the con- 

 nexion between various sections of the colony. In the same 

 ■way the radicle-growths of Idmonea interjuncta are very 

 similar to the cross bars forming the network, whereas in /. 

 Milneana both are stouter ; and this will be referred to when 

 dealing with the Cyclostomata. 



I was surprised to find that no description of these radicles 

 is given in a large number of species where they occur, and 

 also came upon some interesting cases of articulation which 

 had been overlooked, for instance, in Gaherea lata, from 

 Holborn Island ; I have a specimen in which where the 

 branches divide there are two chitinous tubes, which join in 

 a quasi-ganglionic knot, from which a tube is given off to 

 each branch (see fig. 4). 



In thinking this all over it seemed somewhat curious that 

 such a form as Membranipora rohorata should show no indi- 

 cations of articulation, and in consequence I reexamined my 

 mounted specimens to see how far this was the case, and was 

 not surprised to find that in the unilaminate form where a 

 dichotomization had taken place, and where the zoarium is 



