12 THE BKITISH CONULARLE. 



I'elseneer, therefore, has given up entirely the use of the term " Pteropoda " ' 

 iu zoological nomenclature, and has included the three families of " Thecosomatous 

 Pteropods," with a number of other Gastropod families, in his Tribe Bullomorpha, 

 Sub-order Tectibranchiata, Order Opisthobranchiata. 



This consignment of the Pteropods to the Opisthobranch Gastropods (them- 

 selves highly specialised members of the class) makes it impossible to regard them 

 as a primitive group represented in the earliest Palaeozoic rocks. It appears, 

 therefore, that on zoological evidence the inclusion of Gonularia among the 

 Pteropods must be abandoned. 



But the conclusion that Gonularia is not a Pteropod, does little to facilitate the 

 assignment to its natural position. In considering the question attention may be 

 drawn to the following characters of the shell, which appear to be beyond doubt : 



(1) The shell in many cases reached a length of between 100 and 200 mm. 



(2) The shell was in most cases very thin, and usually consisted of at least two 

 distinct layers. 



(3) In at least one species the apical end of the shell was open, and organic 

 attachment to foreign bodies was thus effected. 



(4) In the larger number of species the apical part of the shell was divided up 

 by thin imperforate septa, and the sharply pointed apex was very probably cast 

 during lifetime. 



(5) The aperture of the shell was partially closed by infolding lobes. 



These peculiar and varied characters seem to indicate that Gonularia is a 

 Mollusc, but do not point to a close relationship with any of the main groups. It 

 appears to resemble the primitive Cephalopods more closely than any other forms, 

 and perhaps we should follow von Ihering in placing the genus in this group. 

 But the differences between it and the earliest Orthoceratidae are so great that I 

 should prefer to regard Gonularia as a member of an extinct group, equivalent to 

 the Cephalopods, and derived with them from the same simple shelled ancestor. 



CLASSIFICATION. 



I have adopted Holm's method of classification as being useful, although perhaps 

 it is not of much value phylogenetically. The method is not altogether satisfac- 

 tory, as in following it some species which are probably closely allied appear in 

 altogether different sections. For instance, G. elongata, in the broad, rounded, and 

 horizontal transverse ridges, resembles G. linnarssoni, but by virtue of the sharper 

 definition of these and the fine longitudinal ornamentation it must be placed with 

 the ( ■ancellatae, not with Lmves. Again, no hard and fast line can be drawn between 

 the Moniliferse, and the Cancellatae, G. crassa, G. plicata, etc., forming a transition 

 beween the two. Holm's section, Longitudinales, is unrepresented in British rocks. 

 1 E. R. Lankester, ' Treatise mi Zoology,' pt. v, Mollusca by Pelseueer (1906). 



