76 Cumings — Development of some Paleozoic Bryozoa. 



carinse, and represent a further elaboration of the latter for 

 protective purposes. 



It is still too early to attempt a phylogenetic classification of 

 the Ectoprocta. The present studies will, it is believed, indi- 

 cate the direction in which such a classification must be sought. 

 The budding of all colonial organisms, after the primary stages* 

 is apt to be very irregular, since the growing colony as it in- 

 creases in size is more and more compelled to accommodate 

 itself to the limitations of space and food supply. Only the 

 nepiastic stages, therefore, have any phylogenetic value or any 

 classificatory value higher than generic. The uniformity of 

 the nepiastic stage in Bryozoa (Ectoprocta) suggests a com- 

 mon ancestor, which propagated by means of lateral buds, 

 forming bifurcating series.* Stomatopora presents this type 

 of budding, rarely producing a median bud in addition to the 

 two lateral ones. Where there is a linear series of buds, the 

 writer is disposed to consider the individuals of such a series 

 as representing unpaired lateral buds (fig. 81). Some speci- 

 mens of Siomatopora very forcibly suggest this interpretation. 



Through Probosoina (fig. 82) and related forms, the genetic 

 line from Jin ear and anastomosing zoaria like Stomatopora to 

 true incrusting forms is very complete ; and the development 

 of Fenestella seems to show how cylindrical and infundibular 

 zoaria may be derived from incrusting zoaria. The ancestors 

 of the Ectoprocta (Cyclostomata, Cryptostomata, Chilostomata) 

 may be sought, therefore, in the direction of Stomatopo?*a-\ike 

 types. 



Further than supplying a method, the present studies have 

 done little to clear up the mystery of the Trepostomata. In 

 the writer's opinion, they indicate a rather more remote rela- 

 tionship between this and other orders of Bryozoa than certain 

 authors have held. The Trepostomata are a very ancient type, 

 but they probably do not stand in a linear relation to any 

 other order of Bryozoa.f 



The author's studies strengthen the view advanced by 

 Ulrich 45 that the Cryptostomata are the Paleozoic representa- 

 tives of the Chilostomata. Polypora and Retepora are pre- 

 cisely alike in their early budding stages. 



November, 1903. 



* Davenport 7 holds a somewhat different view. 



f Phylloporina is a composite genus in no way ancestral to Fenestella. 

 P. corticosa belongs to the Trepostomata. 



