INTRODITCTIOX. 
XXI 
But this was not so. I have endeavoured throughout to confine 
the term ‘ cancellus ’ to spaces of interzooecial origin ; hut in some 
genera the cancell i remain as large spaces ; in others they persist 
as long cylindrical spaces forming tuhuli ; while in others they are 
only pores left, like the lunules of some Echinoids, by overgrowth 
of calcareous tissue. In the last case the cancelli may he so small 
and abundant that they render the walls porous or cancellous. 
In 1896 I regarded cancelli as of two main types: long cylin- 
drical tuhuli in the epizoarium, as in the Horneridie, and the simple 
round spaces or spots for which the term ‘ maculae ’ was suggested 
in 1892,* and which are typically represented in the Petaloporidae. 
Fio. 1 . — Hornera lichenoides (Linn.). Fig. 2 . — Petalopora costata (d’Orb.). 
In both cases the cancelli appear to be extra-zooecial in origin, 
though when surrounded by the overgrowth of the thickening walls 
of the zoarium they appear ultimately zooecial. The origin and 
relations of the two types of cancelli may be illustrated by 
Figs. 1 and 2, reprinted from Vol. I of this Catalogue. Fig. 1 
shows the tuhuli of Hornera (the ‘pore-tubes’ of Waters) traversing 
the outer layers of the zoarium. These pore-tubes were, of course, 
occupied by “ protoplasmic and cellular contents,” to use Waters’ 
statement ; these soft materials were no doubt zooecial in origin. 
* J. W. Gregory. “ On the British Paleogene Bryozoa”: Trans. Zool. Soc. 
Yol. xiii. 1893, p. 221. 
Recent : Norway. Longitudinal 
section ; x 12. 
Longitudinal section through wall 
with maculae. (After Pocta.) 
