12S BRITISH STROMATOPOROIDS. 



tubes at much more regular and less distant intervals than it would normally do ; 

 and at the same time to abandon its natural creeping habit, and to send up vertical 

 tubes which continue their growth upwards to an apparently almost indefinite 

 extent. Moreover, instead of producing horizontal stolons at a single level only, 

 namely, in the plane of the general creeping expansion, it must be supposed to go 

 on producing horizontal processes or connecting-tubes at successive levels in the 

 mass. It is, in fact, not uncommon in some types, such as Stromatoporella 

 (Diapora) laminata, Barg., to find such horizontal stolons developed on the upper 

 surface of the last-formed layer of the Stromatoporoid (PI. X, fig. 3), in which 

 cases the appearances produced often closely resemble those presented by an ordi- 

 nary Aulopora-colonj . Professor Ferdinand Roemer has explained the apparent 

 continued growth upwards of the " Gaunopora-tubes " as being perhaps due to the 

 fact that a single " Gaunopora " may be the result of the combined growth of one 

 Stromatoporoid with many successive colonies of Aulopora. I am, however, satis- 

 fied that, in the case of most laminar examples of " Gaunopora " at any rate, only 

 one Aulopora-colonj is concerned, and that the tubes which arise from the basal 

 reticulation are continued upwards through the mass to the upper surface. I 

 believe that this is also commonly the case in the massive examples of " Cauno- 

 pora," though in the case of these it is difficult to prove this positively. 



There are, no doubt, great difficulties in the way of accepting the view that 

 Auloporce when living commensally with Stromatoporoids so fundamentally change 

 their natural mode of growth, as they must be supposed to do if we are to regard 

 them as giving rise to " Caunoporm " and " Diaporce." Upon the whole, however, 

 I think the difficulties in the way of this hypothesis are not so great as those are 

 which confront us if we select Syringopora as the commensal of " Caunopora" 

 Possibly some of these difficulties might be evaded by supposing that in some 

 " Caunoporce" and " Diaporw" the tubes belong to Aulopora, while in others they 

 belong to Syringopora. If we retain the theory of the commensalism of " Gauno- 

 pora," but do not accept either Aulopora or Syringopora as the source of the 

 " tubes," we are driven to the exceedingly improbable hypothesis that these struc- 

 tures belong to a genus of Corals, the species of which are totally unknown, save 

 when living as commensals with certain Stromatoporoids. 



I may just add here that I have found a single example of a " Diapora " from 

 the Devonian Rocks of Devonshire in which the tubes resemble neither Syringo- 

 pora nor Aulopora, but are more like those of the Auloporoid genus Bomingeria, 

 Nich. In this singular specimen, the tubes are aggregated into cylindrical bundles, 

 which would closely resemble the stems of a slender Facliypora, except that they 

 give out at intervals detached tubes which radiate outwards to a considerable 

 distance from the central bundle of tubes. I shall describe and figure this speci- 

 men later on, and need not say more about it at this moment. 



