FORAMINIFERA OF THE ATLANTIC OCEAN 89 



In crushing specimens which had a series of dark brown chambers 

 in the rotaUform portion, I found a thin, nearly transparent brown 

 lining that could be separated from the interior of these chambers, 

 after which they were white. This lining seemed structureless, sug- 

 gesting a chitinous character, and appeared to be the source of the 

 color in the chambers. The specimens had been dried and the lining 

 was of imiform character and thickness, and could scarcely have been 

 dried protoplasm. It may be suggested that a similar condition 

 should be looked for in other colored Rotaliidae, especially where the 

 early chambers only are colored. 



When the "float-chamber" was removed the under side of the rotali- 

 form portion had an appearance very dift'erent from even the small 

 specimens of Cymbalopora poeyi. The suggestion that Tretomphalus 

 hulloides is a stage of Cymbalopora, or of Planorbulina, or of Discorbis,- 

 does not seem to hold in the Tortugas region. All the pelagic speci- 

 mens of Tretomphalus bulloides were much smaller than those of 

 Cymbalopora poeyi, and had a different shape and general character. 

 The ventral side does resemble that of Cymbalopora poeyi, but only 

 in a general way. It may be that these small forms are all megalo- 

 spheric and are thus producing the sexual zoospores for the formation 

 of the microspheric generation, but no specimens referable to the 

 latter were found in the tow-nets during my stay at the Tortugas. 



A peculiar character of the pores of the test was noted. The early 

 chambers have comparatively large pores, which increase in numbers 

 but decrease in size in the succeeding chambers until in the "balloon- 

 chamber" they are very fine, except for the large pores at the base. 



Families 37, 38, and 39 are already discussed in part 7. 



Family 40. GLOBOROTALIIDAE 



Test in the early stages trochoid, the chambers with a rough, 

 cancellated exterior and often spinose, in the adult resuming the 

 ancestral rotalid form or becoming annular, but often retaining the 

 rough, spinose surface; aperture typically opening into the umbilical 

 area, the older species often retaining the covering above the umbil- 

 ical area, and traces of it appear in the hving forms; largely pelagic. 



This family evidently represents a return to the rotalid ancestry 

 of Globigerina, that of a Discorbis-Yike test. 



The compressed forms seen in many of the species of this family 

 have been included in the Globigerinidae, and some of those which 

 have been referred to Pulvinulina in the literature do not fit at all 

 the forms now included under Eponides. The pelagic habit of many 

 of these species, and their association with the Globigerinidae in 

 both Recent and fossil Globigerina marls and oozes, show their close 

 relationship. By reversion to the ancestral form this group helps to 



