244 Geological Society. 



species. Accordingly a study was made of the variation in height 

 presented by the shells, which include every gradation between 

 perfectly discoid forms and types with a spire the height of which 

 considerably exceeds the diameter of the base. By sorting the whole 

 of the shells into ten grades, according to height, it was shown 

 that forms of mean height were common, while extreme forms 

 were rare, the height being distributed, in fact, according to a 

 typical ' variation-curve.' If more than one species were really 

 present, it is in the highest degree improbable that the various 

 types should be distributed in the proportions actually found, and 

 this is taken as the most satisfactory proof possible of the specific 

 unity of the group. 



It is shown that the shells also vary extensively in respect of the 

 amount of carination, the degree of involution, the form of cross- 

 section of the whorls, the form of aperture and the stage of develop- 

 ment at which various chai'acters are acquired, the variation in each 

 character being, however, ' continuous.' The ontogeny of the 

 various characters is considered, and the species is shown to be 

 highly variable at an early stage, the mature characters being 

 largely independent of early variations. 



A discussion is given of the bearing of the inquiry on the 

 stratigraphical application of palaeontological data. 



2. ' The Structure and Relationships of the Carbonicolce' 

 By Miss M. Colley March, M.Sc. 



The evidence for the relationship of the Carbonicolce to the Unionidse, 

 based on shell-structure, muscle-scars, form, habitat, ligament, and 

 hinge-teeth appears insufficient. The first five of these characteristics 

 are shared with obviously unrelated groups. With regard to the 

 last — the teeth — as seen in developed specimens and reconstruc- 

 tions from sections, they seem to be absent. The hinge-apparatus 

 appears to consist of a cardinal plateau, grooved for the reception 

 of an internal ligament. The hinge-plate and ligamental groove 

 are absent in Carbonicola antiqua, very poorly represented in C. 

 turgida and similar forms, and most highly developed in C. aquilina. 

 Another fact which argues against the relationship of the Unionidae 

 to the Carbonicolce is the absence of ornament in the latter group and 

 its presence in the former. This holds good whether the ornament 

 is considered to be due to the effect of the glochidial hooks in the 

 young shells, or to be the remains of ornament : because in the 

 former case it implies the absence of the glochidial stage, and in 

 the latter it implies descent from an unornamented ancestry. 



The position of the Carbonicolce appears to be unsettled, and to 

 be possibly quite different from that of any of the Pelecypoda yet 

 studied : because, according to Bernard's work, the cardinal plateau 

 is developed subsequently to the hinge-teeth, while in the Carbonicolce 

 it is acquired independently of them, specialization taking place in 

 the ligament, for the reception of which the plateau is developed 

 and specialized. 



