134 



UNITED STATES NATIONAL MUSEUM BULLETIN 215 



the inner structures, although any trace of external 

 feature then becomes temporarily concealed. It is 

 therefore difficult to make a comparative examination 

 between external sculpture and internal arrangement of 

 the chambers by this method. 



In studying the internal structm'es (inner characters of 

 the wall, colimaeUar process, toothplate, cribrate or 

 radiate featm^e of the apertiu-e) the best results were 

 obtained by dissection by use of dilute hydrochloric 

 acid mixed with a small quantity of gum tragacanth 

 glue (a method used and described by Troelsen) . This 

 method avoids a dangerous extension of the dissolution 

 of the test as may happen when diluted acid is used 

 alone on very tiny tests. Some of the specimens here 

 illustrated represent dissections obtained by this 

 method, which in many cases can be substituted ad- 

 vantageously for the use of thin sections, and this has 

 made possible many corrections to previous structm-al 

 interpretations. 



Statistical method was only occasionally applied, 

 for it is hardly applicable in many cases, due to the 

 small size of the specimens and the lack of measurable 

 elements. It was used in the investigation of the genus 

 Guemhelina, in order to establish the percentage of 

 coiled specimens in the different species and so to evalu- 

 ate the validity of that genus in comparison with 

 Heterohelix. For this purpose, more than 3,000 speci- 

 mens were statistically examined. 



Systematic Relationships 



With regard to previous interpretations of the rela- 

 tionships in the Heterohelicidae, Cushman (1927a, p. 

 59) described the family Heterohelicidae as follows: 

 "Test in the more primitive forms planospiral in the 

 young, later becoming biserial, in the more specialized 

 genera the spiral stage and even the biserial stage may 

 be wanting and the relationships shown by other char- 

 acters; wall calcareous, perforate, ornamentation in 

 higher genera bilaterally symmetrical; aperture when 

 simple, usually large for the size of the test, without 

 teeth, in some forms with apertural neck and phialine 

 lip." With a range of variability as great as thus 

 stated, almost every perforate foraminifer could be 

 included. In contrast with this too wide allowance of 

 systematic variability for the family, very subtle 

 generic distinctions were accepted between very closely 

 related forms, such as Heterohelix and Guemhelina, 

 which were placed by Cushman in two different sub- 

 famihes because of a distinct early coil in the first and 

 less frequent early coiling in the latter. 



Galloway (1933, p. 342) notes with some humor that 

 "It would be possible to consider the whole group as one 

 without subfamilies, or to make nearly as many sub- 

 families as there are genera, depending upon the cap- 

 rices of the systematist." But some of these genera 

 are quite unrelated. The positions of Psevdouvigerina 

 and Siphogenerinoides were corrected by Galloway, but 

 no substantial changes to the general arrangement of 

 the family were suggested. 



Glaessner (1936, p. 126) divided the Heterohelici- 

 dae, sensu stricto, into two subfamilies: the Hetero- 

 helicinae, containing Heterohelix and Spiropledoides, 

 and the Giimbelininae, including Oumbelina, Gumbe- 

 litria, Tubitextularia, and Psevdotextularia. 



Later, Glaessner (1945, p. 86) observed: "A few 

 families such as the Heterohelicidae and Cassidulinidae 

 are artificial as they include genera whose structural 

 and genetic atfnities lie elsewhere." He separated 

 some of the Heterohelicidae of Cushman into two 

 different superfamilies: Rotaliidea (in which he placed 

 the Giimbelinidae near the Globigerinidae and Hant- 

 keninidae) and Buluninidea, family Buliminidae (in 

 which he placed the subfamilies Bolivininae, Plecto- 

 frondiculariinae and Uvigerininae) . In this publica- 

 tion he used the family name Gumbelinidae, in place 

 of Heterohelicidae. Glaessner's subdivision was the 

 greatest advance to date in the systematics of the so- 

 called Heterohelicidae, for unrelated forms were here 

 definitely separated from the globular-chambered 

 forms related to Heterohelix. 



Sigal (1952) and Pokorn;p- (1954) followed Glaessner's 

 classification in general, both these authors place the 

 family Gumbelinidae (with Guembelina) in the super- 

 family RotaUidea, and place the family Heterohelici- 

 dae (with Heterohelix) in the superfamily Buliminidea. 

 They continued to interpret Bolivinita, Bolivinitella, 

 and Bolivinella as an homogeneous group within the 

 Heterohelicidae. 



The recent tentative classification of a group of 

 Heterohelicidae from the Upper Cretaceous of the 

 Pyrenees, made by Kikoine (1948), is based upon such 

 erroneous interpretations as the biseriality of Guem- 

 belina. Moreover, Kikoine considered only six genera, 

 leaving undiscussed the trio Bolivinita, Bolivinoides, 

 and Bolivinella, and he failed to discuss their most 

 important characters. 



No systematic rearrangement is possible without a 

 previous revision of the genera on the basis of their 

 type species. In this connection some recent contri- 

 butions must be mentioned. Loeblich (1951) empha- 

 sized and illustrated the presence of coiling in "Gum- 

 belina," and "Ventilabrella," and noted the biserial, 

 rather than triserial, initial stage in EouvigerinM. 

 Hofker (1951b) examined the structxire of Bolivinoides 

 and the "toothplate" in Bolivinita, discussing new 

 morphologic elements. Stone (1946) described the 

 inner structure of Siphogenerinoides in comparison 

 with Siphogenerina. 



These few analytical contributions clearly demon- 

 strate the exactness of the statement by Loeblich 

 (1951, p. 106) that "few families among the Forami- 

 nifera contain genera as poorly known as are several 

 genera belonging to the family Heterohelicidae." 



Basis of Present Revision 



The following variable elements have been considered 

 in this study: (1) CoUing in the early stage; (2) shape 

 of the test and arrangement of chambers in neanic and 

 adult stage (acceleration, etc.) ; (3) position and shape 



