290 H. M. Brydone — Furthev Notes on the Trimmingham Chalk. 



and other perishable bodies must have been fairly plentiful in the 

 Chalk sea— ^iu the zone of B. quadrata we, find great numbers of free 

 small Serpnla (nm^ullacea&n^granulatn) whose bases show ol early that 

 they grew attached to some long slender body, no doubt a seaweed — 

 and any specimens which had grown attached to such perishable bodies 

 would now be indistinguishable from those which had never been 

 encrusting. These two forms of growth are, in fact, not unfrequently 

 shared by Cretaceous species, e.g., Homalostega vespertilio and 

 Crihrilina Gregoryi, though the majority of encrusting forms are 

 practically never found free, and some free forms, e.g. Semieschara 

 Camii (post), are never found encrusting. If D'Orbigny's unilamellate 

 genera, which are separated only by tliis point of growth, be mei-ged, 

 we have left a number of genera based on habits of growth which 

 constitute for the Cretaceous forms convenient and often natural groups 

 which do not seriously overlap. No doubt they are artificial, but 

 that is not yet a destructive criticism of any classification of the 

 Polyzoa. 



But though it may well i-emain a matter of individual opinion 

 whether habits of growth are of any value in the definition of 

 genera, I do not think that any field student of the Cretaceous 

 Cheilostomata would deny them great value in the definition of 

 species. It is very rare in the English Chalk for any species to 

 occur in more than one habit of growth (treating the free and 

 encrusting unilamellate forms of growth as identical). A better 

 test could hardly be ofi'ered than the two very similar Onycliocellas 

 figured respectively by D'Orbigny as Celleporn Paristensts and by 

 Beissel as Eschara galeata (the latter species is the one commonly 

 identified as E. Lamarcki, but which corresponds only to Beissel's 

 figure of E. yaleata). E. galeata is a bilamellate form and the most 

 abundant Cheilostome at all horizons above the zone of J/, cor- 

 testudinarinm. The specimens which have passed through my 

 hands must run well into tens of thousands, yet I have onl}" one 

 encrusting specimen and no mature free unilamellate specimen, 

 notwithstanding that it commonly grows from an unilamellate base. 

 Cellepora Parisiensis, on the other hand, is one of the commonest 

 encrusting species, and I must have seen thousands of specimens, 

 yet I have no bilamellate specimen (and only one free unilamellate 

 specimen). Stronger testimony to the fidelity of these species at 

 any rate to a certain habit of growth can hardly be possible, and 

 nearly all species of Cretaceous Cheilostomata show equal or nearly 

 equal fidelity. Here, then, we have a point of considerable 

 importance, in which the rigid application to Cretaceous forms of 

 canons of classification derived from the study of Tertiary and 

 recent forms is very undesirable. 



Another such point is to be found in the prevailing tendency to 

 unite Cretaceous, Tertiary, and recent specimens in one species. 

 This involves the general assumption that polyps which produce 

 skeletons which cannot be distinguished must be identical in 

 organisation. We can admit the absolute propriety of this assump- 

 tion in the case of specimens more or less contemporaneous and 3'et 



