124 
TREPOSTOMATA. 
as in the Cyclostomata, and (2) that they bend suddenly outward 
along part of their course and then change in character. The 
sudden bend is not confined to the Trepostomata, for we see it 
in such Cyclostomata as Liversaria tuhiporacea and Sparsicavea 
carantina, d’Orb., of which illustrations of thin sections, reprinted 
from Yolume I, are shown in Figs. 30 and 31 ; while in various 
Trepostomata the zooecia gradually increase in size without any 
sudden change in character or direction, as is illustrated by 
Fig. 32, p. 131, Fig. 42, p. 162, and Fig. 45, p. 166. 
In well-preserved Mesozoic Trepostomata, e.g. Multicrescis tuherosa 
(Fig. 54, p. 207), the change in the character of the zooecia after 
the distal bend is well marked ; the walls become thicker and 
moniliform, as in the typical Palaeozoic forms. The moniliform 
aspect of the walls is doubtless due to their having been pierced 
by pores or canaliculi. Slight solution of the walls, such as often 
occurs during fossilization, would readily convert the canaliculi 
into funnel-shaped openings separated by bead-like walls. 
Kicholson’s account of the structure in the recent New Zealand 
species of Heteropora shows that in that species the zooecia 
undergo a sudden change at the distal end, that the distal walls 
are pierced by canaliculi, that the zooecia have diaphragms or 
‘tabulae’ and radial spines, and that the large zooecia are about 
*25 mm. in diameter. In all these respects this Heteropora agrees 
with the Monticuliporoids, although all these characters are not 
present in all the genera of that group and the spines are so 
delicate that they are rarely preserved. 
Nicholson, however, regarded the above eharacters as merely 
superficial resemblances, and some of them, such as the presence 
of the diaphragms, may be explained as independent homoplastic 
developments ; but the value of this group of characters seems 
to me greater than the differences which led Nicholson to refer 
MonticuUpora to the Coelenterates and Heteropora to the Bryozoa. 
The three differences on which he relied are that MonticuUpora 
has imperforate walls, no radiating spines, and a different 
structure in the different types of corallites ’ in dimorphic or 
trimorphic species. But these characters are not valid. In regard 
to the first, TJlrich includes ^ in his diagnosis of the family 
* E. 0. Ulricli. ‘‘Palaeontology of Illinois,” sect. vi. Palaeozoic Bryozoa: 
Geol. Surv. Illinois, 1890, vol. viii. p. 369. 
