BRYOZOA FROM ZANZIBAR. 8353 
highly differentiated Cheilostomata that we must at present look 
for guidance, rather than to the simpler Cyclostomata. 
There are, however, some characters in the Cyclostomata, 
which when more used will help to show which forms are nearly 
velated. Smitt showed that the ovicells furnished important 
characters, and on several occasions I have pointed out the 
importance of the ovicells, and of what I called the peristome of 
the ovicellular opening, saying in 1888* that ‘“ the position and 
nature of the opening is most important, eften more so than the 
shape of the ovicell.” © Harmer has confirmed what I said, and in 
several very valuable papers has gone into detail in some genera, 
and has named the passage by which the larva escapes the 
occiostome, while the external orifice is the oceciopore, which 
terms are generally accepted, though unless new terms are 
absolutely necessary they are always to be regretted, as every 
branch of science 1s now overloaded with names. Also the size of 
the embryo, and of the zocecial aperture, as well as the position 
and nature of the closures, give specific or generic characters. 
The primary zoecium and the growth of the younger part of 
the zoarium must receive much more attention, in fact up to the 
present in recent forms it has received hardly any: for example, 
in Hntalophora proboscidea, E. deflera Couch, and Filisparsa tubu- 
losa there is a Stomatopora-like growth often spreading for a 
considerable distance over the supporting material, before the 
zoarium becomes erect, whereas in what has been considered to 
be Entalophora rugosa VOrb., from Naples, there is a Diastopora- 
growth often covering a considerable piece of the stone or shell 
upon which it grows before the erect cylindrical zoarium is 
formed. It is elsewhere shown that other characters, namely 
the lamina and ovicells, prove that it belongs to the Diastoporide. 
To return to the more highly differentiated Cheilostomata : in 
various genera the zoarium may be either adnate, erect, uni- or 
bilaminate, or even articulated. For example, Lepralia, Schizo- 
porella, Celleporidee, all occur adnate, erect, uni- and bilaminate, 
while Zhalamoporella may be unilaminate, bilaminate, adnate or 
erect and articulated ; Cellaride in recent forms are usually 
articulate, except in the younger branches, but in fossils fre- 
quently there is no articulation. Now in none of these cases is 
there any material difference in the zoecium according to the 
way in which the zoarium grows. 
It is sometimes forgotten that all this only deals with the 
position of the zoeecia, or, as Levinsen would say, the “‘autozooids,” 
and this has proved of quite secondary value in classification ; but 
this does not mean that characters furnished by the other 
‘‘zooids” such as avicularia, stalks, stolons, radicles ete., which 
also cause differences in the form of the zoarium, may not furnish 
valuable characters. 
No doubt the Catenicellide are derived from unarticulated 
* “On some Ovicells of Cyclostomatous Bryozoa,’ Jour. Linn, Soc., Zool. 
vol. xx. p. 276. 
