292 



are here somewhat larger and that only a small number of them bear avictilaria. 

 It seems reasonable to conclude therefore, that this division into areas is here 

 also a sign, that this covering is composed of kenozooecia. Waters, who has 

 extended our knowledge of the Bryozoa on so manj^ points by bringing forward 

 new or insufficiently known structural features, has also at several places made 

 investigations on this characteristic division of the covering into areas. Thus, in 

 his paper * on a part of the Bryozoa material of the Challenger Expedition he has 

 called attention to the fact, not only that it shows stratification but also that it 

 contains numerous cavities. In a later paper ^ he has further remarked, that by 

 incineration it can be loosened from the true zooecia, and that the lines which 

 divide it into areas are not merely superficial marks but that they extend right 

 through it. As the incineration however makes this covering unsuitable for closer 

 investigation, and boiling in caustic potash, by which means two-layered colonies 

 can usually be divided into their two layers, here leads to no result, there seems 

 no other way of isolating this kenozooecial layer than to grind down the zooe- 

 cial layer, and this I have done with good results in the case of Ret. tesselata. 

 A fragment of Ret. lata, which was sent me by the British Museum, shows quite 

 exceptionally here and there a tendency of this layer to scale in small plates, 

 and these thus offer favourable conditions for a closer investigation. After such 

 a plate has been boiled for some time in caustic potash, we can by means of a 

 needle isolate the single kenozooecia, which thus possess not only independent 

 lateral walls, but also, what is never found in the ordinarj^ zocecia in any Bryo- 

 zoa with exception of the Onychocella species, independent distal and proximal 

 walls. I have found the same thing in the kenozooecia of Ret. tesselata and it will 

 probably prove to be the rule in tliis family. The above-mentioned kenozooecia 

 of Ret. lata also appear to contain an inn«r cavity, which however is of an ex- 

 tremely variable and irregular form and consists, e. g. in the two connected keno- 

 zooecia figured on PI. XXIII, fig. 2 a, of a number of small, more or less elongate 

 cavities connected by narrow, canal-like parts. The cavities in adjacent kenozooecia 

 are also connected in the same way, and the separating walls thus show a corre- 

 sponding number of round openings or rosette-plates. PI. XXIII, (figs. 2 b, 2 c) 

 shows the same kenozooecium seen from the basal (zooecial) surface and from 

 the one side. In the first case it shows a large, inner cavity, which however is 

 in parts interrupted by calcified portions and is not seen with the same distinct- 

 ness everywhere, as it lies at a somewhat different height at different places. It 

 is in connection both on the distal wall and on the two lateral walls with the 



1 110, pp. 19—21, figs. 7,9. 2 115, pp. 77—78. 



