TROPICAL PACIFIC FORAMINIFERA OF "ALBATROSS" 35 



collections, but it seems to be well distributed in the deeper samples 

 and is present in several variable forms. 



In its most typical form this species is distinguishable from A. 

 madagascariensis by its flatter and more equally biconvex test, by its 

 angled rather than curved sutures, and by the greater distinctness 

 of its sutures owing to the limbate ridges of clear shell material and 

 to the addition of glassy knobs between the sutures. 



The sutures extend out almost radially or even slightly forward 

 for about half the distance from umbilicus to periphery, then angle 

 sharply backward at about a 90° angle and continue backward until 

 they merge into the periphery. The inner ends of the sutures usually 

 are irregular in pattern whereas the outer parts form a smooth curve 

 (pi. 14, fig. 3). As a result of this suture pattern, characteristic of 

 the species, some of the individuals (pi. 14, figs. 1-3) take on the 

 appearance of being partially evolute, particularly on their ventral 

 sides. The evolute appearance of some specimens (pi. 14, fig. 2) may 

 be attributed partly to the transparency of the shell material with 

 which each whorl covers the previous whorls. 



The supplementary chambers that extend outward from the cen- 

 tral area on the ventral side are shaped quite differently in A. radiata 

 from the way they are in A. madagascariensis. In A. radiata the 

 peripheral extensions of the supplementary chambers are so con- 

 stricted in most specimens that they appear as little more than 

 thickened suture lines, and it may be difficult to determine just how 

 far out the supplementary chambers extend. 



The umbilical areas on both dorsal and ventral sides are occupied 

 by a more or less conspicuous and raised glassy knob. In many 

 individuals this glassy knob serves as a window through which the 

 interior of the test may be viewed (pi. 14, fig. 3). In rare individuals 

 the knob is so large and high it stands up as a transverse stud pro- 

 jecting from the center of the test (pi. 13, fig. 1). 



Degree of ornamentation of the surface of the test varies from 

 smooth to densely papillate. In the more heavily ornamented species, 

 chambers and sutures are totally obscured. Number of chambers 

 per adult whorl ranges from about 9 to 15. Flatness of the test is 

 likewise a variable feature. 



Some of these variants are illustrated (pis. 13 and 14). Others 

 have been described and illustrated elsewhere. Some have been 

 given varietal names — var. papillosa Said and var. venosa (Fichtel 

 and Moll) . It seems likely that these variations are not determined 

 ecologically but rather that the species is naturally variable. 



