291 



seen on figures of Ret. sinuosa Kirkp. S Ret. novae zelandiae Waters^ and Rhyn- 

 chozoon profundum Mac Gill.^ as well as on the accompanying figures of Rhync. angu- 

 latum n. sp. (PI. XXIII, fig. 4 a). In those forms which have a strongly developed 

 peristome, the arch named is difficult to see from the frontal surface and to 

 examine it we must grind down the basal wall of the colony. This also applies 

 to the hinge-teeth. The rosette-plates seem to show great constancy, and in the 

 numerous species I have examined in this regard I have only found two in which 

 the rosette-plates had several pores. Thus, the distal wall in R. lata is provided 

 with a narrow, transversely oval rosette-plate with a row of 3 — 5 pores and in 

 Rhgnc. angulatum some of the rosette-plates may have two pores. The rosette-plates 

 in this species are exceptionally situated at the bottom of pore-chambers. The 

 ooecia also in spite of smaller differences show great agreement in their structure. 

 The peculiar layer of kenozooecia, which covers the basal surface of the 

 free-growing colonies and is of the same nature as the expansion attaching 

 them to their under-layer, deserves closer description. With regard to this expan- 

 sion Smitt* has correctly recognised in Ret. elongata (Ret. Wallichiana Busk), 

 that it consists of imperfectly developed individuals, and he has even found some 

 of them with an aperture. Hincks^ describes these individuals as » aborted cells «, 

 but as I have already shown in my » Studies on Bryozoa« he*' has in so far mis- 

 understood this expansion, that he regards it as the first formed part of the 

 colony, whereas in reality it only arises after a number of ordinary zocecia have 

 been formed. It then gradually increases in extent with the further growth of the 

 colony. PI. X, fig. 1 c shows a beginning colony of R. Beaniana, which shows in 

 addition to a primary zooecium in the Membranipora stage two fully developed 

 zocecia and the basal surfaces of two just beginning. Fig. 1 d on the other hand 

 shows a slightly older colony with a small radical expansion and in figs. 1 g and 

 1 h parts of this are magnified to a greater extent. Both show beginning keno- 

 zooecia on the growing margin the membranous roof of which has disappeared 

 in the previous boiling in caustic potash; their distal wall is provided with 1—2 

 small rosette-plates. If we examine the outer surface of a Retepora colony attached 

 to its under-layei', e. g. the colony of Retepora Beaniana figured on PI. X, 1 a, 

 which is attached by its radical expansion to a tube of Hydroides norvegica, we 

 see that the division into irregular, mostly avicularium-bearing areas shown by 

 this expansion is continued without interruption or boundary on to the outer 

 side of the free part of the colony, with this difference, however, that these areas 



1 49 a, p. 269. ' 108 c, PI, VII, figs. 1-3. ' 108, PI. II, fig. 15. ' 100, p. 200-201, PI. XXVIII, 

 fig. 232. ^ 22, p. 388. " 22, p. 394, fig. 18. 



19* 



