rORAMINIFEEA OF THE KERIMBA AECHIPELAGO. 613 



2 Stations. 



One specimen, large and typical, built up of molluscan fragments and sand-gvains 

 Avith one or two grains of magnetite, at Stn. 1, and another broken specimen lacking 

 the initial spiral portion at Stn. 8, in which the proportion of magnetite grains used in 

 the construction of the shell is large and striking. None of the other Foraminifera at 

 Kerimba building adventitious tests employs magnetite, and, as this mineral is not 

 abundant in the Kerimba sands, its occurrence in these specimens of //. agqlvutinann 

 confirms our statement as to the selective power exhibited by this species. The only 

 Kerimba sand containing any marked proportion of magnetite grains is that from 

 Stn. 13 (PerabaBay), which is geographically at the other extremity of our series of 

 samples. We have discussed this tendency exhibited by H. agglutinans to incorporate 

 magnetite in its shell in several of our papers [cf. H.-A. & E. 1909 TNS. p. 411 

 et passim). 



134. Haplophragmium compressum Goes. (Pi. XL VI. figs. 20, 21.) 



Lituolina irregularis, var. compressa Goes, 1882, RRCS. p. 141, pi. xii. fiu-s. 421-423. 

 Rhaphidohelix elegans Mobius, 1880, FM. p. 76, pi. ii. fig-. 2. 

 Haplophragmium emaciatum Brady, 1884, FC. p. 305, pi. x.xxiii. figs. 26-28. 

 Egger, 1893, FG. p. 262, pi. v. figs. 53, 54. 

 „ compressum Gees, 1896, DOA. p. 31. 



Millett, 1898, etc., FM. 1899, p. 359, pi. v. fig. 8. 



1 Station. 



A number of specimens at Stn. 13, characterized by very light o-rey, almost white 

 tests, built of sand-grains firmly cemented together. Spicules entirely absent. T!ie 

 septation is very obscure. The test is umbilicate on both sides, and the final chamber 

 in large specimens often presents a tendency to expose the preceding whorl in an 

 irregular manner. 



The Kerimba specimens cannot be described as strongly marked or typical, but we 

 have little hesitation in assigning them to Goes' species. They difl^er, however from 

 the types of Goes in the entire absence of spicular material. Goes appears to lay 

 considerable stress on the spicular habit of his type, as he refers to the preponderance 

 of spicules in the test. He also compares his type with Eliaphidohelix elegans Mobius 

 and Hajplophragmium foliaceum Brady. There can be no doubt as to the close resem- 

 blance of Goes' type to that of Mobius both in shape and in the preponderance of spicular 

 material, but Brady's species bears no very close resemblance to the type of Goes and 

 moreover, H. foliaceum never in our experience utilises spicules for the construction of 

 its test. It seems possible therefore that Goes intended to refer to Hajjlophragmium 

 emaciatum Brady, which is of similar shape to Goes' type, and, moreover utilises 

 spicules largely in the construction of its test. We have accordingly referred 

 II. emaciatum Brady to the earlier type of Goes, while leaving II. foliaceum apart. Too 

 much importance must not, however, be placed on the utilisation of spicular material 



