ROTALIINAE 417 



whorl rising at a sharp angle from the carina ; at about half their width they are flattened 

 out, and so continue to the inner margin. At the point of curvature there is a second 

 carina, projecting upwards and running parallel to the peripheral edge. The sutural lines, 

 which are flush between the two carinae, become limbate between the secondary carina 

 and the inner margin, so that the inner half of each chamber is surrounded by a raised 

 wall. Within these walls the surface of the test is rough, thus obscuring the earlier con- 

 volutions, which can, however, be made out on careful examination. Between the inner 

 and outer carinae the shell surface is smooth and hyaline. 



Owing to superficial ornament, only the chambers of the final convolution are 

 exposed on the ventral side, which is flat and somewhat depressed at the umbilical 

 region. This part of the test is thickly studded with minute beads of shell-substance. 

 The aperture is obscure, but is believed to be a minute sHt on the inner ventral side of 

 the final chamber opening into the umbilical region. 



Two or three specimens only were found at WS 88, and a single one at WS 245, both 

 stations lying between Tierra del Fuego and the Falkland Islands. The structure will be 

 most readily understood from the figures (figs. 25-27). The two specimens from WS 88 

 appear to be megalospheric ; the nature of the third cannot be stated. The dimensions 

 of the largest test are: greatest breadth, 0-30 mm.; least breadth, 0-25 mm. ; thickness, 

 0-09 mm. 



A single large specimen from WS 88 (greatest diameter, 0-40 mm.), which we also 

 figure (figs. 28-30), may be abnormal, or it may be a further stage in the development of 

 the species. The peripheral carina is much reduced and the inner carina has disappeared. 

 The ventral area of the dorsal side is covered with oval pits surrounded by raised edges. 

 Each pit apparently marks a chamber. From the pits a series of radiating grooves 

 extends to the carina. The ventral side is very similar to our Type but the beaded orna- 

 ment is only slightly developed. 



The exact relationships of this remarkable organism must await the discovery of 

 further material, as it is quite unlike anything with which we are acquainted. It has 

 been tentatively assigned to Discorbis mainly on the evidence of its beaded base, but it 

 has a general likeness to some Truncatulinae, such as T. temdmargo and T. altocamerata, 

 and might equally well have been assigned to that genus. 



Genus Heronallenia, Chapman and Parr, 193 1 



354. Heronallenia (Discorbis) kempii(Heron-AllenandEarland)(Plate XVII, figs. 20-28). 

 Discorbis kciiipii, Heron-Allen and Earland, 1929, etc., FSA, 1929, p. 332, pi. iv, figs. 40-8. 

 Heronallenia kempii, Chapman and Parr, 193 1, NAP, p. 236, pi. ix, figs. 6-8. ' 



Eleven stations: 48, 388; WS 84, 86, 87, 88, 89, 91, 92, 93, 248 (see also p. 309). 



Test free, perforate, white in colour, consisting of flattened chambers arranged in a 

 rapidly expanding coil of at most i\ convolutions. On the dorsal side, which is rather 

 convex, the marginal edges of the chambers are strongly limbate, and the whole surface 

 between these limbations is decorated with exogenous beads and zig-zag ornament which 

 conceal the arrangement of the earliest chambers. There appear to be about eight chambers 



