WEAVER: INVERTEBRATE PALEONTOLOGY AND HISTORICAL GEOLOGY 699 



basis of this interpretation Neumayr built up a phylogenetic classification based 

 in part on the' stratigraphic range of species and genera. 



Ludwick Rhumbler (1899) considered that the evolutionary development of 

 foraminifera during geological time involved morphological changes which would 

 produce a strengthening of the test and that the uniserial forms would be more 

 easily injured or destroyed than the coiled types. Accordingly, he regarded the 

 former as ancestral to the more complexly coiled forms, a conclusion which is not 

 entirely in agreement with the information concerning the occurrences in the 

 fossil record. 



The evolution of the foraminifera was discussed in an important publication 

 by Henri Douville (1907). In this he considered that the composition of the 

 test was dependent to a large extent on the differences of environment under 

 which the animal lived, that the calcareous forms were descended from the are- 

 naceous, and that those with uniserial tests represented a more advanced state 

 of development than the coiled forms. At present the arenaceous types are not 

 considered ancestral to the calcareous. The classifications presented by Cushman 

 from 1927 to 1940 raised the number of families to 49, which Cushman believed 

 were derived largely from coiled, nonseptate genera; he also thought that the 

 calcareous group had arenaceous ancestors. His classification accepts many of 

 the principles advocated by Douville. Galloway in 1933 criticized Cushman's 

 classification, asserting that the latter had established too many families and 

 subfamilies and that many of the genera had been placed in wrong families and 

 in incorrect relationships between families. Galloway's classification and ar- 

 rangement of families were based on his interpretation of the phylogenetic char- 

 acter of the different groups and determined from a consideration of comparative 

 morphology, the stratigraphic range of the different assemblages of genera, and 

 the biogenetic law. He considered the foraminifera as consisting of two broad 

 groups, one of which evolved from the Allogromiidae, the other from Endothyra. 

 The former constitutes a group of foraminifera with a single chamber and, be- 

 cause of its chitinous wall, is rarely preserved as a fossil. However, he consid- 

 ered it ancestral to the nonseptate calcareous and arenaceous forms, including 

 the Miliolidae. 



The most recent classification is by M. F. Glaessner, published in 1945. The 

 order Foraminifera is divided into 7 superfamilies, 37 families, and 38 subfami- 

 lies. Glaessner concludes: 



1. The non-septate forms are more primitive than the septate. 



2. The higher, or septate, spirally coiled arenaceous foraminifera form a well-defined 

 group. 



3. The Fusulinidae are derived directly from Endothyridae. 



4. The different lines of the porcellaneous foraminifera have a common origin in a 

 coiled non-septate form. 



5. The Polymorphinidae are derived from Legenidae but there is no clear evidence 

 concerning the origin of this family. 



6. The Cassidulinidae and Ellipsoidinidae (Pleurostomellidae) are related to the 

 Buliminidae which can be traced back to a trochospiral ancestral form. Most of the other 

 smaller calcareous perforate foraminifera are clearly derived from rotaloid (trochospiral) 

 ancestors. 



7. Most of the larger calcareous perforate foraminifera including Siderolites, Orbi- 

 toides, Lepidocyclina, Miogypsina, and probably the nummulites, developed from a number 

 of different but closely inter-related small rotaloid (trochospiral) ancestors. 



