an introduction to the biology and systematics of komokiacea

AN INTRODUCTIONTO THE BIOLOGY ANDSYSTEMATICS OF KOMOKIACEABYOLE S . TENDAL and ROBERT R . HESSLERZoological LaboratoryUniversity of CopenhagenScripps Institution of OceanographyUniversity of CaliforniaABSTRACTKomokiacea n . superfam . (Textulariina. Foramini- 1928. Lana n . gen., Baculella n . gen., and Edgertonia n .ferida. Protozoa) comprises agglutinating foramini- gen . are described . Two families. Komokiidae andfers with test consisting of a complex system of fine. Baculellidae. are erected .branching tubules of even diameter . The test wall is The group has worldwide distribution. mainly insimple. with argillaceous particles . Stercomata accu- the abyssal zone . The greatest relative abundancesmulate within the tubules .have been found in abyssal oligotrophic areas and inEleven new species in the genera Komokia n . gen., hadal trenches .Septum n . gen., kjoa n . gen., Normnina Cushman.I Introduction and acknowledgements ........................... 166I1 Materials ................................................................... 167111 Methods .................................................................... 170IV Morphological partA . External appearance of the kornoki ......................... !7!B . The tube wall ......................................................... 172C . Pores ..................................................................... 172D . Colour and consistency ......................................... 172E . The interior of tubules ............................................ 173F . The pseudopodia ................................................... 173V Geographical distribution ........................................... 174A . Horizontal distribution ........................................... 174B . Vertical distribution ................................................ 175VI Position in deep-sea communities ................................ 175VII Systematic partKomokiacea in the literature ....................................... 177Proposed classification of the Komokiacea .................. 177Descriptions ............................................................... 177Superfamily Komokiacea n . superfam .................. 177Key to the families and genera ............................. 178Family Komokiidae n . fam ........................... 178CONTENTSGenus Komokia n . gen .............................. 178Komokia multiramosa n . sp ..................... 179Genus Septuma n . gen ............................... 179Septum ocotillo n . sp ............................. 180Genus $0," n . gen ..................................... 180Ipoafragila n . sp .................................... 181Genus Normanina Cushman, 1928 ............. 181Normnnina tylota n . sp ............................ 182Normanina saidovae n . sp ........................ 184Genus Lana n . gen .................................... 185Lana neglecta n . sp ................................. 186Luiii reticulata n . sp ............................... 186Family Baculellidae n . fam ........................... 187Genus Baculella n . gen .............................. 187Baculella globofea n . sp ......................... 187Baaclella hirsuta n . sp ............................. 188Genus Edgertonia n . gen ............................ 190Edgertonia tolerans n . sp ......................... 190Edgertonia argiilispherula n . sp ................ 190VIII Summary ................................................................... 192IX References ................................................................. 193

I. INTRODUCTIONIn the more than 100 years since deep-sea investigationbegan in earnest on the "Porcupine" and the"Lightning", the expeditions sent out from manynations have collected enormous quantities of materialfrom the deep-sea floor. These investigationshave employed a wide variety of gear, includingcameras, dredges, trawls, grabs, corers, and submersibles.Many specialists covering a span of severalgenerations have been involved in the workingup and the publication of the results in hundreds ofbooks, reports and papers.The basic composition of the deep-sea fauna wasrevealed fairly early in this history. There are virtuallyno important groups that cannot be found inthe "Challenger" collections. However, some majortaxa have turned up from time to time in recentyears. Notable were the discoveries of the Pogonophora(Johansson 1937, Ivanov 1963, NQrrevang1970) and the monoplacophoran Neopilina (Eemche1957, Lemche & Wingstrand 1959). Nevertheless,while these are major taxa, they are rarely or neverdominant components in deep-sea communities.Thus, it would seem safe to maintain that it is highlyunlikely that any important macrofaunal taxon isunrecognized in the literature.Strangely enough, this prediction is wrong. Thereexists in the deep sea an essentially undescribedmacrofaunal taxon that can be found in virtuallyevery sampIe where one looks for it. Furthermore,in samples from the oligotrophic abyss, the bulk ofthis taxon exceeds by far that of all the metazoans puttogether. Not only is this group abundant, but it isdiverse. Single stations contain a wide variety of specieswhose morphologies differ in ways that are oftengross, reflecting several higher taxa. Many of thesetaxa are cosmopolitan in distribution.The animais in question are smali (only a few miliimetersacross) clusters of minute branching tubules.They are fragile and often obscure because of thesediment which covers their surfaces. We knowfrom experience that they are often ignored or discardedfram general co!!ections during the sortingprocedure because they look like lint, nondescriptorganic detritus, poorly-washed balls of sediment, orminor fragments from the surface of some undefinedlarger organism. In cases where they might be recognizedas benthic organisms, they could reasonablyonly have been considered sponges or foraminifers.However, sponge specialists reject them as lackingany of the features of that phylum, and they are sounlike the morphologies classically regarded as belongingto the Foraminifera that specialists of thatgroup are loath to accept them. Hessler (1974) andHessler & Jumars (1974) have suggested that theymay belong to the Xenophyophoria, but Tendalrejects this possibility. Thus, for the most part thefew specimens that have been retained rest in limbo.The only major programs in which these animalshave been retained by routine are those of Hesslerfrom his samples in the abyssal Pacific, the morerecent samples of the Soviet Union's Institute ofOceanology, and the French investigations in the Bayof Biscay. The Russians call these animals "vetvistyekomoktchki", which means "branching lump" (A.Kuznetsov, personal communication). This appellationseems especially appropriate, and therefore weshall continue to call them "komoki" (singular"komok") as a matter of convenience.As a result of our investigations, we conclude thatkomoki are indeed foraminifers of the agglutinatingtype. However, except for their being members ofthe suborder Textulariina, they are for the mostpart completely unacceptable by the present taxonomicberarchy. We therefore erect the newuperfamily Komokiacea to include them.The problem with describing a group such as thisfor the first time is that it is so vast. Single samplescontain dozens of species, and we suspect that thetotal number of species in just our few hadal, abyssal,and bathyal samples may be in the hundreds. Clearly,it will take years of laborious effort to bring theknowledge of komoki up to the level of that of otherrecent taxa. We therefore choose in this first treatmentto confine ourselves to describing this superfamilyand some of the lower taxa which typify therange of morphologies that clearly belong within it.In liming ourse~ves in this way, we Ft!!jr recognizethat there will be many important morphologies thatwill need future treatment.We will, furthermore, discuss in the Iight of presentinformation the distribution of komoki and theirposition in deep-sea communities.AcknowledgementsWe are happy to extend our thanks to the manypersons who aided us in this study: J. Lipps (Universityof California, Davis), R. Douglas (University ofSouthern California), H. Thiel (University ofHamburg), Z. A. Filatova (Institute of Oceanology),

and C. G. Adams (British Museum of Natural History)for useful discussion and information; C.Edgerton, B. Burnett, E. Flentye, R. La Borde (all ofUniversity of California, San Diego), 6. Ulstrup, andL. Hoffmann (both of University of Copenhagen) forpreparation of specimens and manuscript; K.Elsmann (University of Copenhagen) for drawings,and G. Brovad and P. Jeppesen (both of University ofCopenhagen) for other technical aid, Special thanksgo to F. J. Madsen (University of Copenhagen) forhis criticism of our manuscript. The following institutionswere kind enough to lend us Komokiacea fromtheir collections: British Museum of Natural History(London), Centre OcCanologique (Brest), Institutfiir Hydrobiologie und Fischereiwissenschaft(Hamburg), Institute of Oceanology (Moscow),Scripps Institution of Oceanography (La Jolla),Smithsonian Institution (Washington), and ZoologicalMuseum (Copenhagen). This work was supported bythe Danish Natural Science Research Council(51 1-5437), the United Stated National ScienceFoundation (GB 14488, GA 3 1344K, DES 74-21506),and the United States Energy Research and DevelopmentAdministration (AT (04-3)-34 PA21 3).Our samples come from several sources. The core ofthe collection is part of Hessler's sampling of the abyssaloligotrophic community under the central NorthPacific gyre (Hessler 1974, Hessler & Jumars 1974).We have concentrated our efforts on this material,firstly because we are familiar with the way in whichit was collected and with the community from whchit comes. Secondly, because we then can illustrate therange of morphologies that live together. Thirdly,because we have box core samples from this areawhich will ultimately allow quantification of the organismson the bottom.The samples used for comparative morphologicalstudy and for determination of geographic rangehave been selected from collections borrowed byTendal from institutions in other countries (Table 1).Table 1. Collections from which supplementary samples have been selected.Expeditions Geographical region No. of samples Depository of collection"Valorous", 1875"Galathean, 1950-1952N. Atlantic OceanRound the world1 British Museum of Natural History,London8 Zoological Museum, Copenhagen"Vitiaz", 1958,1959, 1969 N. Pacfic Ocean 23 Institute of Oceanology, MoscowInternational Indian Ocean Expedition,"Anton Bruun", 1964"Meteor" 3,1966Ckiii~ XI:, "Ago", 1969W. Indian OceanIberian SeaX. Pacific Ocean1 Smithsonian Institution, Washington3 Institut fiir Hydrobiologie undFischereiwissenschaft, Hamburg4 Scripps Institution of Oceanography,La Jolla"Eastward", 1972 W. Atlantic Ocean 3 Zoological Museum, CopenhagenSouth Tow - leg 1,"Thomas Washington", 1972Cruise 14, "Akademik Kurchatov",1973"Thalassa", 1973Biogas VI, "Jean Charcot", 1974Eurydice - leg 7,"Thomas Washington", 1975Wingstrand, 1975Wingstrand, "Dana", 1976Equatorial and SouthPacZicCaribbeanBay of BiscayBay of BiscayPhilippine TrenchKorsfjorden, NorwaySkagerrak, E. North Sea2 Scripps Institution of Oceanography,La Jolla39 Institute of Ocea~?o!agy, Mosco?~17 Centre Oct5anologique, Brest10 Centre Octanologique, Brest2 Scripps Institution of Oceanography,La Jolla2 Zoological Museum, Copenhagen1 Zoological Museum, Copenhagen

Table 2. List of the stations from which the samples of Komokiaceaused in the present investigation originate.AB - Anton Bmun; Ar - Argo; Bi - Biogas; Eastw - Eastward; Gal - Galathea; Ku - Akademik Kurchatov; Me - Meteor; Tha -Thalassa; Va - Valorous; Vi - Vitiaz; Wash - Thomas Washington; Wing - Wingstrand.AD - anchor dredge; BC - 0.25 mZ box corer; BD - Bodlot dredge; BT - beam trawl; ES - epibenthic sled; OG - okean grab; OT -otter trawl; PG - 0.2mZ Petersen grab.Atl. - Atlantic; Carib. - Caribbean; Pac. - Pacific.eutr. - eutrophic; olig. - oligotrophic; sublitt. - sublittoral; trans. - transitional.Ship Stat. No. Depth (m) Latitude Longitude Location GearABAr---Bi---Eastw-Gal-----Ku-W. Indianeutr., abyssalCent. N. Pac.olig., abyssalCent. N. Pac.eutr., abyssalCent. N. Pac.eutr., abyssalN. N. Pac.eutr., hadalE. Atl.eutr., abyssalE. Atl.eutr., abyssalE. Atl.eutr., abyssalE. Atl.eutr., bathyalW. Atl.eutr., abyssalW. Atl.eutr., abyssalE. Eq. Atl.eutr., abyssalE. Eq. Atl.eutr., abyssdW. Indianeutr., abyssalW. Indianeutr., abyssalE. eq. Pac.eutr., abyssalE. eq. Pac.eutr., abyssalCarib.eutr., hadalCarib.eutr., abyssal

Ship Stat. No. Depth (m) Latitude Longitude Location Gear- -Carib.eutr., abyssalCarib.eutr., hadalCarib.eutr., hadalCarib.eutr., abyssalCarib.eutr., bathyalCarib.eutr., bathyalE. Atl.eutr., abyssalE. Atl.eutr., bathyalE. Atl.eutr., bathyalE. Atl.eutr., bathyalN. W. N. Atl.eutr., abyssalS. E. N. Pac.trans., abyssalN. E. N. Pac.eutr., sublitt.N. E. N. Pac.eutr., bathyalN. E. N. Pac.eutr., abyssalN. E. N. Pac.eutr., bathyalCent. N. Pac.oiig., abyssalCent. N. Pac.oiig., abyssalCent. N. Pac.oiig., abyssalE. Eq. Pac.eutr., abyssalCent. S. Pac.olig., abyssalW Eq Pac.eutr., hadalW. Eq. Pac.eutr., hadalE. Atl.eutr., bathyalE. Atl.eutr., bathyal

At ths early stage in our understanding of the tax- reasonable confidence. The samples used for thisonomy we have not attempted to determine the ge- purpose are from the stations listed in Table 2.ographic range of our species. On the other hand, The holotypes are deposited in the Zoologicalthe distribution of genera can be discerned with Museum, Copenhagen.One of the important reasons why komoki have notbeen recognized in the past is that the techniquesused for handling deep-sea material have in largemeasure been inadequate. The following commentsare intended to serve as a guide to handling materialin the search for and investigation of this taxon.HandlingKomoki do not have a rigid test; they are often softand always fragile. Thus, samples must be handledwith care. Preferably there should still be sedimentin the sample when it reaches the sea surface; thissediment will insulate the animals from mechanicaldamage in the water column.The sample should be washed gently. The elutriationtechnique of Sanders, Hessler & Hampson(1965) is recommended. Since komoki are oftensmall, a fine-mesh screen should be used; a 300-500pm mesh would be adequate.FixationThe washed sample must be fned immediately inorder to reduce as much as possible cytologicalchanges after death. Hessler's samples, which are themain part of the present material, were fined in buffered4 % formaldehydelsea water solution and after2 days transferred to 80 % ethanol for storage.Although this procedure is commonly accepted asusable for small marine animals, it often poses difficultiesfor histological workers because of shrinkaged~ring stages fo!lowing fiixatior?. In nxr samples thefollowing features may possibly be signs of such aneffect: the crooking of tubules in some species, thewavy appearance of the inner organic layer of thetest in places, the extremely diffuse condition of theplasma. and the strong tendency of the stercomata tobe arranged in tight clumps filing only part of thetubule lumen.Concerning more detailed cytological and cytochemicalwork in future we recommend that tests beperformed with furation in Bouin's fluid, Carnoy orClarke, Flemming's fluid, Susa and Zenker.Because of their fragility, komoki cannot be driedwithout special procedures.StainingRose bengal, the stain commonly used to discriminateliving foraminifers, does not work well on komoki. Amore promising stain seems to be the mixture ofeosin B and biebrich scarlet proposed by Williams(1974). It should be used in a 1: 1000 concentration ofeither 4-5 % formaldehyde as given by Williams, ordistilled water. Alcohol does not work at all, or atbest extremely faintly on protozoans. It should benoted that there are differences between genera(and species) of komoki as to the degree they arestained: Septuma, Lana, Baculella, and Edgertoniastained differently, but strongly in about 12 hours,whereas Komokia, Ipoa, and Normanina stained faintlyor not at all, even when staining was prolonged to 96hours. Destaining can be done with 96 % alcohol or4-5 % formaldehyde. Alcohol is the most effective,removing practically all stain in 48 hours.SEM preparationThe procedure used for preparation of specimensfor scanning electron microscopy was critical pointdrying (Nemanic 1972). The technique was modifiedfor use on the Komokiacea as follows: a 6 mm holewas punched in the lid of a No. 00 Beam capsule, anda 50 pm plankton netting was inserted into the capover the hole. Solutions were injected with a hypodermicsyringe via a No. 26 needle through the otherend of the capsule.From storage in 80 % ethanol, the komoki werehydrated, immersed in a 2 % solution of AgNO, for10 minutes at room temperature and rinsed in threechanges of distilled water. They were subsequentlytreated with Cajal's reducing solution (Luna 1968)for 5 minutes, and rinsed again in two changes ofdestilled water.Dehydration was followed by two extra changes of100 % ethanol for 15 minutes each, a 1:l solution ofethanol and freon 113 for 15 minutes, and finally twosolution changes with freon 113 for 15 minutes each.The freon 113 was drained from the capsule, whichwas quickly placed within the freon critical pointdryer (Cohen et al. 1968) and dried using freon 13.The specimens were affmed to an aluminum stub

with silver conducting paste and coated undervacuum with approximately 2WA of gold-palladium.They were viewed with a Cambridge S-4 Stereoscanscanning electron microscope.SectioningConsiderable information about morphology can belearned by examination of paraffin sections, of whichwe recommend 5-7 pm as a useful thickness. Schiff sPAS reaction, hematoxylin, Heidenhains Azan, andtoluidine blue are general stains which highlight especiallythe inner organic layer of the test. Toluidineblue has the disadvantage that it may stain the outeragglutinated layer too heavily.The terminology used is the same as for Foraminifera.The size range given of a single species must bejudged in relation to its morphology, because it issometimes difficult to distinguish between small, entirespecimens and fragments of larger ones.One of the aims of this paper is to show the exceedinglylarge morphological variation found within theKomokiacea. A person unfamiliar with the groupmight therefore come to the conclusion that the descriptionof species is easily done, requiring only afew discrete characters. Although it may be so insome genera, in general this will not be the case. Ourexperience is that a number of morphologies stronglyresembling each other, when closely examinedmust be judged as representing separate species. Thisis the case both within single samples and betweenwidely distributed samples.It is desirable to avoid the state of affairs knownfrom several other taxa rich in species that a greatmany descriptions are incomplete in important charactersand inadequate for later work. Therefore,when describing a new species of Komokiacea oneshould always pay attention to the following features,although the elucidation of some of them may berather time-consuming:1. Growth form and outer dimensions.2. Details of branching, including anastomosingversus nonanastomosing.3. Presence or absence of beads.4. Diameter of tubules.5. Presence or absence of sediment filling in interstices.6. Type of agglutinated material.7. Structure of the tube wall, with dimensions.8. Presence or absence of chambering (can only bereliably worked out through sectioning).9. Presence (and structure) or absence of foramina.10. Distribution of stercomata.11. Appearance of plasma and nuclei.A. Extemd appearance of the komokiKomoki are visible to the naked eye, their largestdimension generally being 1-5 mm.A tremendous number of different, characteristicbody forms are to be found endowing thistaxon with a degree of diversity which at leasteijuak that seen in the other two superfamilies ofthe suborder Textulariina, and which is comparableto that found e. g., among hydrozoansand bryozoans.The basic element in the morphology ofkomoki is a fine tubule of even diameter, rnntainingthe plasma and a large quantity of faecalmaterial, generally in the form of pellets calledstercomata. Different organizational patterns ofthe tubules are reflected as different bodyforms. The external appearance is, moreover,strongly influenced in some species by the inclusionof sediment in the interstices betweentubules.IV. MORPHOLOGICAL PARTThe general body form of komoki is of twosorts. There may be a center of organization asin trees or bushes; here one may speak of centraland peripheral tubules or at least of a centerof growth (Pls 9A-D, 10, 11A-D). In the other casethere is no center of growth or organization, justas wouid be the case with a baii of fleece or ascouring pad (Pls 13, 14A-B, 16C).Branching is the basic feature controllingbody form. In most cases branching is dichotomous,although tri- or polychotomy also exists(PIS !()a, 14e-D). Stems and I.-'.-ui ail~lll~l~ ---7-- L ~ C VC~Lcomprise more than one tubule. Diameter andcross section of the branches are often constantthroughout the whole tubule system. There are,however, a number of cases where they arefound to be different when comparing a stem(central tubule) to its side branches, and sometimesthere is stepwise reduction following eachlevei of branching. The ends of tubules are some-

times bulbous (Pls 11E-F, 12E) but most oftenstraight and rounded (Pls 10, 11A-B).Two types of branching have turned out to beof constant occurrence. In the first, the tubulesare of cylindrical form, with only rare constrictions,and the distance between succeedingbranching points is much larger than the branchdiameter (Pls 10, 14C-D). In the second type,numerous, usually close-set, very short sidebranches give the tubule a beadlike appearance(Pls 15B, 16E). Here, constrictions of the maintubules are often seen, and the short sidebranches are often constricted at the base (PI.16A-B, D, E). While branching of the first typegenerally seems to follow a simple pattern, in thesecond type it is quite often complicated by tightlyrepeated branching and partial lateral fusion.Anastomosing is a feature found in a fairlylarge number of species. We limit this designationto mean that direct communication betweenthe cavities of two tubules occurs. It does notapply to the mere fusion of the outer agglutinatinglayer of two tubules which have grown contact.B. The tube wallIn the tube wall two layers can be distinguished insections seen with the compound microscope (Pls21B-D, 22, 23, 24). The inner layer is generallyvery thin, less than 0.5 pm thick. It looks smoothon the interior surface, and stains readily kithtoluidine blue or PAS. Scanning electron micrographsdemonstrate that the inner layer is frequentlylaminated. On the outside this layer issharply demarcated from an agglutinated layerof sediment into which it may in some speciesproject long fibers.The agglutinated layer varies considerably inthickness, but 10 ym can be given as frequentmeasurement. The included particles are commonlyof the clay and silt fractions, although--&:-I-- -C ---A &OP O-P fn~lnrl i2 the tests cf~LLILILIGJ Ul 3QllU JWU ULC. L V U l l Uspecies from comparatively shallow localities.The organic material which holds the sedimenttogether stains with toluidine blue or PAS and insome species shows reaction for organicallybound iron when treated after Pearl's or Turnbull'smethods.In some species, long flexible tubules projectfrom the general surface of the body (Fig. 10,p.191). These tubules differ from the rest inhaving a very thin outer, agglutinated layer.Several species have been found to be septatewith the chambers added in a linear arrangement(Pls 21A, 24A) and sometimes with conespondingfaint constrictions on the surface of thetubules (Pl. 12A). Septae are two-layered, consistingof the inner organic layers of two successivechambers (Fig. 4; Pls 21B, 24A). They neverinclude parts of the agglutinated layer. Somesections show that the inner organic layer may belaminated and that a very thin outer lamina mayfollow the outer agglutinated layer, the resttaking part in the formation of the septum. Secondarilyformed, simple foramina which can beprovided with a circular, thickened neck exist insome species (Pls 21A, 24A).6. PoresKomoki are devoid of real apertures. Thepseudopodia must penetrate the tube wallthrough minute pores, but possibly these are ofonly temporary character.In the surface of some specimens very smallopenings can be discerned with a dissectingmicroscope. The SEM photos of the exteriorsurface of the tubules show a wide size range ofsmall (most of them

tral mass which is round or strandlike, dependingon their arrangement.The inner organic layer and the organiccement in the agglutinated layer provide theanimal with a rather high degree of flexibility. Ifthe agglutinated particles are few or small thereseems to be space enough between them to allowfor a certain degree of bending. If they are numerousor of sand size, there is a correspondingloss of flexibility.E. The hterior of hbulesPlasma and stercomata are found in the tubulelumen (Pis 21B-D, 22B, 25A). Sections and closeinspection with the dissection microscope revealthat in many species a large part of the tubulesystem is empty (Pl. 26C-D).The plasma is difficult to stain. It may either bestrongly diluted or have the character of finethreads, or it may be so filled with ingested materialin vacuoles or lacunae that it is hard to see.In a few cases plasma is encountered in theSEM photos (Pls 17B, 18A-B). Although in someareas it looks homogenous, the main impressionis that it is organized in an extensive network ofvery fine strings. There seem to be no inclusions,but commonly plasma strings were found in intimatecontact with the surface of stercomata ofthe rugged type mentioned below.Bodies resembling nuclei have been found inhistological sections of most species. They arespherical to ovoid, with granular interior (Pl.21B, D). The size variation ranges over an intervalof about 5 pm in any one species, being fromabout 4 to 10 pm in diameter for the wholegroup. For some reason the Feulgen reactionbib iiot work; these bodies resemble, however,nuclei after staining with haematoxylin-eosin andHeidenhains Azan. They can be distinguishedfrom stercomata in unstained sections. In a singlespecies, nucleoluslike bodies with diameter about2 pm were seen.The most conspicuous elements in the tubules arethe stercomata or faecal pellets (Pls 21B, 22B, 23B-C,25A). They are opaque or dark brown, generally upto about 10 pm in diameter, although they have beenfound as large as 35 pm, and are composed of a massof unidentifiable small particles held together by asubstance of the acid mucopolysaccharide type. Withthe compound microscope they look smooth on thesurface, in the SEM photographs some look smooth,others rugged (Pl. 19E-F).The phases in the process of formation of stercomataare expected to start with the uptake by thepseudopodia of particles from the environment.These are accumulated inside the test as loose massesof "ingested material" (PI. 21C). During the digestiveprocesses they are transformed into stercomatawhich finally are left in the tubule system. Reasonsfor this retention could be either that some phase inthe digestive process is dependent on the accumulationof particles in large masses which are too largeto be passed back out throught the very smallopenings in the test wall, or that the stercomata areof some use, for instance as a weight to counteractbuoyancy. Another possibility is that the stercomataafter some time in the tubule lumen are recolonizedby microorganisms and can be redigested by thekomokiacean. There is no indication of this, but if itexisted it would be of the uttermost importance.Morphological specialization and accumulations offaecal pellets in the hindgut of the abyssal Abraprofundorumhave been proposed as an arrangementallowing bacterial growth with subsequent redigestionof the faecal material (Allen & Sanders 1966). Asomewhat similar nutritional pattern is possible to acertain extent for Hydrobia (Newel1 1965) and isattributed a considerable role in some decomposerfood chains (Fenchel 1972).F. The pseudopodiaOn one occasion during the scanning process, partsof a branch of Septuma ocotiIlo were seen to becovered by a formation consisting of two elements inclose contact (Pl. 20A-B). One is a coherent, relativelyhomogenous, although heavily perforated layer (Pl.2OC-Ej. The other is seen as a reticuiation of undividedor dichotomously divided strings which aregenerally of even thickness (0.1-0.2 pm in diameter)and most often straight (Pl. 20E). Some strings connectirregular lumps of various sizes (0.5-2 pm indiameter), while other strings contact a lump at oneend and either seem to penetrate the underlyinghomogenous layer, or possibly to join it with theirother end.The whole formation may be what is left of a partof the outstretched pseudopodial system. Beforecapture, it may simply have covered the test surface,or it may have been extended into the surroundingmud/water.

The homogenous, perforated layer resembles thepseudopod reticulum adjacent to the test as seen inIridia diaphana after freeze drying (Marszalek 1969).One can imagine a number of interpretations ofthe string-lump element. It should, however, benoted that what makes the surface of some stercomatarugged, is the presence of irregular lumps ofan appearance and a sue variation comparable tothat of the lumps found on the presumed pseudopodia1system (Pl. 20C). Moreover, this type of stercomatais generally found in contact with what seemsto be plasma strings (Pis 18B, 19E).Thus, what has been described above can be inter-preted as part of the pseudopodial system showingphases of transport of nutrients. The pattern of foodtreatment could be much as reported in Allogromialaticollaris, where nutritive particles are transportedby the pseudopodia as "lumps" into the test lumen tobe digested, not in vacuoles, but in extracellularspaces, lacunae (Lengsfeld 1969). Differences fromthe Allogromia type of food uptake would be thatkomoki take in a large number of very small particles(selected because of presence of microorganismson the surface?), digest them in the tubulelumen, and then leave the indigestible parts as faecalpellets inside the test.V. GEOGRAPNICAL DISTRIBUTIONFrom the present collections, comprising more than100 samples, we have selected about 40 to elucidatethe geographical range of the Komokiacea (Table 2).Not all our samples were included, because in anumber of areas there were stations lying close toeach other. The criteria for selection were such as toallow the greatest attainable horizontal and verticalcoverage. Only taxonomic levels down to genus areconsidered, because only a few samples (from thearea of the central North Pacific gyre, Hessler 1974,Hessler & Jumars 1974) have as yet been treated tospecies.A. Horizontal dis&ibutionThe superfamily Komokiacea has been found in thethree main oceans and very near to the Arctic(Table 2 and Fig. 1). This implies a cosmopolitan distribution,the obviously large gaps in the distributionbeing due to lack of investigation, poor methods ofwashing, and misidentifications.~oth families are represented in the three mainoceans.Four of the seven genera, viz., Septuma, Ipoa,Lana, and Baculella, are distributed in the Atlantic,the Pacific, and the Indian oceans, while Komokia,Fig. 1. Distribution of samples in which Komokiacea have been found.174

Normanina, and Edgertonia have been found in theAtlantic and the Pacific.KomokiidaeBaculellidaeB. Vertical distributionThe vertical range of the samples in which we havefound Komokiacea is 400 to 9600 m, with by far thelargest number of stations in the depth interval from1000 to 7000 m (Fig. 2).Although (as Fig. 1 shows) our samples are farfrom being evenly distributed over the deep-sea bottomof the world ocean (we have particularly largeconcentrations of stations in the North Pacific, theCaribbean area, and in the Bay of Biscay), on thebasis of available information it seems reasonable toconclude that the Momokiacea is essentially a deepseagroup. Both in the Atlantic and in the Pacific wehave several samples taken between 500 and 1500 m,indicating that the group has its shallowest distributionsomewhere on the continental slope, perhaps thelimit being physically correlated with the shift fromsandy to clayey silt and with strongly diminishedtemperature variation.One record is as shallow as 180 m, but this may notbe reliable. There is only one specimen, and it mayhave been left in the gear from a previous deepersampling.The two families seem to have about the samedepth distribution in the abyssal and hadal zones (Fig.2). The Baculellidae has not been recorded shallowerthan about 2000 m.Septuma and Lana are represented at true bathyaldepths, while Normanina and Baculella appear between2000 to 3000 m. Komokia has only been takenfrom the lower abyssal zone, whereas Zpoa andEdgertonia also appear in the abyssal but extend welldown into the hadal zone, together with eurybathicLana and Baculella. Septuma and Normanina extenddown into the lower abyssal.Fig. 2. The vertical distribution of families and generaof Komokiacea.Because nearly all macrofaunal animals in the deepsea are very small (Sanders et al. 1965), screens witha mesh opening of 0.3-0.5 mm must be used toachieve complete retention of this faunal category.Careful sorting of samples washed through suchscreens reveals large numbers of komoki.The greatest relative abundances of komoki havebeen seen in the oligotrophic abyss and in hadaltrenches, as exemplified by samples from the centralNorth Pacific (Hessler & Jumars 1974) and from thePhilippine Trench. In such cases, their volume greatlyexceeds that of all the metazoans combined andequals that of the remains of other Foraminifera.The relative abundance of komoki seems to be less in

the equatorial abyss and in the bathyal zone, as seenin the eastern equatorial Pacific and Southern Galiforniacontinental borderland, respectively. Thesedata suggest that komoki are often a dominant, or atleast an appreciable element in deep benthic communities.It is difficult to quantify such relationships in termswhich are meaningful for community analysis.Because komoki are so fragile, individuals areusually represented in a sample by several fragments.Thus, simple counts are insufficient to documenttheir abundance. Obviously, at this early stageof investigation the samples are too valuable to beutilized for biomass measurements (wet weight, dryweight, organic carbon, etc.), since such techniquesare more or less destructive.Because of the test, especially in cases where it isagglutinated, it is generally difficult to determinewhich foraminiferan individuals in a sample are alive,unless special procedures are employed to reveal thepresence of plasma. The high percentage of emptytests in shallow water samples is explained by shortlife span of single individuals (all plasma enters intothe production of gametes and agametes), andlorhigh durability of the test.In komoki the reproductive cycle is unknown. Thetest has a comparatively low content of organic matter,the inner organic layer being very thin and theamount of cement being small, which gives reason tobelieve that decomposition time may be short.This suspicion is supported by histological studies.Ten specimens of Septma ocotillo, 5 of Lam neglecta,and 4 of Normim tylota were sectioned and inspectedfor evidence of having been alive at the timeof sampling. The criteria used were the generalcondition of the body and the presence of nucleuslikebodies, especially after staining with WeidenhainsAzan and haematoxylin. On these bases, all individualswere alive. These data must be viewed with caution,because no attempt was made at randomsampling. However, if these observations are typical,then most komoki are alive. Even if true, such anobservation must be interpreted with the understandingthat only a small portion of a komokiantubuie system actually contains iiving piasma.To document the diversity of komoki, we analyzed500 cm2 vegematic subcores (Jumars 1975) of a 0.25m2 box core (H-153) from the type locality in the centralNorth Pacific. Five randomly selected subcoresfrom the same sample yielded 56 species (Table 3).Table 3. Number of species of Komokiacea found in500 cm2 of box core sample H-153. The 56 speciesare roughly assigned to the morphological typesrepresented by the known genera.Morphological typeNo. of speciesLana .....................................................Komokia ...............................................Edgertonia .............................................Normanina ............................................BaculeNa ...............................................Ipoa ......................................................Septwna ................................................Others ..................................................General observations suggest that such high diversityis typical of abyssal samples from this area (Wessler& Jumars 1974), as is reasonable in view of the greatstability of the physical regime.These subcores were partitioned into layers andtherefore allow analysis of the distribution of komokiwithin the bottom (Table 4). More than 80 % of thespecimens were in the top 1 cm while there wereonly 2.5 % below 2 cm. This restriction of most individualsto the most surficial layers conforms to thepattern displayed by the rest of the fauna, and is whatone would expect in a nutrient-poor, physically calmenvironment.Interestingly, some species showed greatest abundancein the 1-2 cm layer, and 5 species (of Lana andBaculella morphologies) were not at all found in the0-1 cm layer, indicating that at least some komoki livebeneath the surface. This observation demonstratesthat caution should be used in deducing komokian lifestyles from morphology and appearance alone.Table 4: The distribution of species and specimens inthe upper iO cm of 5W cm2 of box core sample W-153.Depth (cm) No. of species No. of specimens~otal 50+r6 1984

VPI. SYSTEMATIC PARTKomokiacea in the literatureOnly one genus, Normanina, has been treated in theprevious literature, commonly as a member of thefamily Astrorhizidae.Poorly known organisms consisting of fine tubulescan be found in "incertae sedis" positions in manydifferent phyla. It is recommended that no such speciesbe included in the Komokiacea without inspectionof material, although merely noting the existenceof these "problematica" may be useful (see e. g. p.185).Proposed classX1cation of the KomokiaceaThe system proposed here includes the descriptionof eleven species, all of them new to science. Theseare to be regarded as a necessary formal foundationfor the taxonomic pattern of the superfamily.The decisions have been made on the basis ofmuch wider experience gained through the investigationof a large number of undescribed species.Superfamily Komokiacea n. superfam.1. Family Komokiidae n. fam.Genus Komokia n. gen.Genus Septuma n. gen.Genus Ipoa n. gen.Genus Normanina Cushman, 1928.Genus Lana n. gen.2. Family Baculellidae n. fam.Genus Baculella n. gen.Genus Edgertonia n. gen.DescriptionsThe higher level classification used here is accordingto Honigberg et al. (1964).Phylum ProtozoaSubphylum SarcomastigophoraSuperclass SarcodinaClass RhizopodeaSubclass GranuloreticulosiaOrder ForaminiferidaSuborder TextulariinaRemarks: A crucial point in all diagnoses of theorder Foraminiferida or the subclass Granuloreticulosiais the presence of delicate, finely granularreticulopoda. These have, however, only beendescribed in a few species, and although there seemsto be agreement on the characterization of thepseudopodial system as given above, nearly nothing isknown about details of the structure and movement.The classification of a given organism as a foraminiferis generally done on test characters, and the Komokiaceais no exception to this. The features whichwere decisive for including the Komokiacea in theForaminiferida were the following: an organic("chitinoid") layer covered with a layer of foreignmaterial held together by a secreted organic cement;the test with one to many chambers; accumulationsof stercomata in the interior; pseudopodia supposedto protrude from simple wall perforations.KOMOUCEA n. superfam.Diagnosis: Test consists of complex system of fine,branching tubules of even diameter. Test wall simple;agglutinated particles argillaceous. Stercomata(faecal pellets) accumulate within tubules.Type genus: Komokia n. gen.Remarks: In referring to tubules as fine, we meanthat the ratio of the size of the individual to the diameterof one of its tubules is very high. Referring toammodiscacean tubes as coarse implies the oppositeof this.When more information accumulates, other charactersmay turn out to be of diagnostic value also atthis level, viz., the finer details of tubule wall structure,the apertures, the organization of the plasma,and the stercomata formation and arrangement.Dissussisn:The detailed knowledge of the morphology of manyagglutinating foraminifers is still scarce, a fact thatmakes strict demarcation of subgroups difficult andcomplicates transferring of earlier described speciesfrom one group to another without direct examination.In our materials from many parts of the worldthere is a large number of undescribed agglutinatingforaminiferan species referable to several new genera,which we arrange into two new families. Thestep to erect a new superfamily for these within thevast array of forms of the suborder Textulariinamay seem rather drastic. However, the common

morphological pattern of komoki does not satisfactorilyfit either of the two existing superfamilies.This judgment is based mainly on the classificationof the Foraminifera as outlined by Loeblich andTappan (1964, p. C184 ff.). To allow direct comparisonwith the definition of the Komokiacea we restatethe diagnoses of the two other superfamilies in aslightly modified and rearranged form.Ammodiscacea: Test nonseptate or only irregularlyconstricted, of simple form, irregular, spheroidalor tubular; tubular types branching, straight, orenrolled; tubules generally coarse, often tapering orirregular. Test wall simple or labyrinthic; agglutinatedparticles arenaceous or argillaceous. Aperturesto external environment simple, in tubular formsgenerally large.Eituolacea: Test composed of coarse chamberswhose arrangement is spirally coiled, bi- or triserial,or straight; rarely branching. Test wall simple orlabyrinthic; agglutinated particles arenaceous orargillaceous. Apertures single or multiple.Lituolacea is the best defined of the two, a fact thatappears not only from the diagnosis, but also whencomparing the families included.Ammodiscacea is one of those heterogeneous taxaso well known from many other parts of the animalclassification, consisting of a collection of widely differentgroups which remain when other, well definedtaxa are brought into their proper context.This is one of the reasons why the distinction betweenthe Komokiacea and the other textulariines appearsblurred. The fate of some genera hitherto includedin the subfamily Dendrophryinae (family Astrorhizidae)illustrates the situation: Syringammina wastransferred by Tendal (1972) to a family of its ownwithin the subclass Xenophyophoria, and Normaninais here placed in the new family Komokiidae.Key to the families and generala. Tubules of cylindrical form; distance between succeeding branchings several timestubule diameter ...................................................................................... Fam. Komokiidae 2 (p. 178)lb. Tubules of beadlike appearance because of numerous short side branches; distance betweensucceeding branchings 1-3 tubule diameters ........................................ Fam. Baculeliidae 6 (p. 187)2a. Tubules anastomosing ................................................................................................ Lana (p. 185)2b. Tubules free .................................................................................................................................... 33a. Body organized into base and radially extending branches, each bearinga distal sweliing ..................................................................................................... Normanina (p. 181)3b. Body bushlike; branches without distal swelling ................................................................................ 44a. Tubule diameter markedly decreasing following each branching ....................................... Ipoa (p. 180)4b. Tubule diameter nearly similar throughout individual ........................................................................ 55a. Tubules septate; branching primarily at base ............................................................ Septuma (p. 179)5b. Tubules nonseptate; branching becomes more frequent distally .................................. Komokia (p. 178)6a. Body stick- or club-shaped ...................................................................................... Baculella (p. 187)6b. Body rounded ....................................................................................................... Edgertonia (p. 190)KOMOKIIDAE n. fam.of complexity in overall branching pattern. At present,there is not enough general information on theDiagnosis: Body variously shaped: tree- or bush- Komokiidae to estimate the value of characters suchlike, or tangled clump. Tubules basically cylindrical septation and presence of dista! swe!lings, ,nd ~fin form, infrequently constricted, and witn distance more special features of branching such as markedbetween branchin@ generally longer changes in diameter of tubules following divisions.than diameter.Type genus: Komokia n. gen.Remarks: We prefer to put the genera in an orderreflecting conditions of primary tubules and degreeKomokia n. gen.Diagnosis: Body bushlike, branching mostly dichotomous,nonanastomosing, occurring anyplace along

Fig. 3. Section through part of a tubule ofKomokia multiramosa n. sp.AGGLUTINATEDLAYERPLASMA-@ VARIOUS STAGESIN FORMATIONOF STERCOMATA,AS&=,:@&STE RCOMATAWITH FINEGRAN-ULAR CONTENTStubules, with increasing frequencyery. All tubules of essentially thenonseptate.toward periph-Type species: Komokia multiramosa n. sp.Distribution: Cent.N.Pac.: H-29, H-30, H-36.Carib.: Ku- 1 18 1, Ku-1187.Material: H-30.Komokia multiramosa n. sp.(Fig. 3; Pls 9A-B, 1 1 A-B, 19F)Diagnosis: Same as that of genus.Description: Body formed like a radial bush, butcomprising only sector of sphere, usually hemisphereor less. Primary branching exposed at lowersurface, at center irregular, dichotomous to multiple,with bases of branches often defined by slightconstriction. Thereafter, branching occurs irregularly,although there is a tendency that each successivebranching halves the distance to the tip.Commonly four to six stages of branching, but up toten have been found. Diameter of body up to 2 mm.Tubules near base thicker than those at tip, but notmarkedly so; basal thickness 35-75 pm in diameter;peripheral diameter 35-50 pm. Tubules crooked,but generally trending in single, basically radialdirection. Tips rounded. Tubules flexible, but stiff.Colour greyish owing to visible filling of tubuleswith stercomata. Agglutinated particles clay.Inner organic layer of tubule wall

sepiama ocotidlo n. sp.(Fig. 4; Pls 9C, 10A-B, 12A-B, 19A, 20A-F, 21A-D)Material: H-30.Diagnosis: Same as that of genus.Description: Body formed much like a bush whosemany primary branches radiate irregularly outwardfrom a single area, whose branching may bedichotomous or multiple, but where the interbranchdistances are always very short. All additionalbranching is dichotomous. Secondary branchingmay occur close to base. Tertiary and quaternarybranchings are well separated by long unbranchedzones. Size of body up to 3 mm across. Length frombase to tip up to 2 mm.Tubules which form branches give appearance ofbeing of similar diameter throughout length, althoughthey taper almost imperceptibly; near base50-105 pm in diameter, toward tips 30-55 pm. Tubulescrooked between branchings, but only modestly so,such that on the whole they appear to be straight oronly gently curved. Sides of tubules straight, onlyfaintly constricted at level of internal septa (Pls10A-B, 12A-B). Tubule tips rounded. Tubules flexible,but stiff.Colour tan, darker toward base, probably becauseof thicker accumulation of agglutinated particlesthere. Agglutinated particles clay, or atcoarsest, silt. In distal parts of the branches are consecutivegroups of stercomata, each representingone chamber; from the outside they look like a stringof pearls.Inner organic layer of tubule wall 0.5-1 pm thick.Outer layer of agglutinated sediment 5-20 pm thick,commonly around 10 pm.Cavity of tubules subdivided into nearly equalchambers by regulzirly spaced, transverse septa (PI.21A). Chambers are 30-75 pm long and 20-50 pmacross (Pl. 21B). Generally they are longer thanwide, and a common size is about 50x30 pm. In shortchambers the wall is often wrinkled, possibly as anartifact.Septa are constrictions of inner organic layer,laminae of which pull away from agglutinated layer atthis point (Fig. 4; P1. 21A-B). Consequently, they arein principle double layered. In each septum is asimple, circular foramen which generally is centrallyplaced, although occasionally it may be acentric (PisFig. 4. Diagram ofSeptum inSeptuma ocotillo n. sp.agl - agglutinatedlayer, iol - innerorganic layer,f - foramen,chl -chamber lumen.pm. The ring which is 34 pm thick merges graduallywith the rest of the septum.Chambers more or less filled with stercomata "which are round or ovoid, more rarely irregular,and measure 5- 10 pm in diameter.Nuclei (Pl. 21B, D) are ovoid or spherical,measuring 5- 12x6- 15 pm. They look granulated in thesections, and are, at least in some cases, providedwith a spherical nucleolus measuring about 2 pm indiameter. The distribution of the nuclei is difficult toelucidate because they do not stain well. Commonlyonly one is found in a chamber, but from time to timethere may be two. In a few specimens nuclei werefound in most chambers, but most often they are notseen. No particular pattern could be recognized intheir distribution along the branches.Remarks: On some specimens are rounded, lightformations 100-200 pm in diameter, most often nearthe base and more or Iess entangled with thebranches. These give the impression of being a basalswelling of the body of S. ocotillo. However, the sectionsreveal that these are in fact specimens of anagglutinating rhizopod, using S. ocotillo as substratum.It is distinguished from S. ocotillo by thefollowing: no inner organic layer in the wall, nochambering, several nuclei present in the undividedplasma, plasma filled with small inclusions, devoid ofstercomata, particles of the test coarser, and wallthickness generally much larger.Ipoa n. gen.19A, 21A). Edge of foramen thickened, forming ring Diagnosis: Major fragments treelike with basal stemwith inner diameter of 8-12 pm, in a single case 18 followed by burst of tightly spaced multiple branch-fchl

ing. Diameter of tubules decreases markedly witheach branching. Tubules nonseptate.Type species: Ipoa fragila n. sp.Distribution: Cent.N.Pac.: H-29, H-30.E.Atl.: Bi-DS78.Carib.: Ku-1181, Ku-1187, Ku-1189,Ku- 1 194.W. Indian: AB-382C.Remarks: The branching pattern in combinationwith the strong decrease in tubule diameter createsan appearance much like that of segments of a headof cauliflower.Consistency stiff, although somewhat flexible.Colour of most of the body tan, but faintly greyish indistal branches owing to barely visible core of stercomatain the tubule cavity. Agglutinated particlesclay.Inner organic layer of tube wall exceedingly thin(~0.5 pm), outer layer of agglutinated particles 5-15pm thick, with thicknesses of 5 pm restricted to distalbranches.Cavity of tubules nonseptate. In central tubule andprimary branches it is nearly empty, whereas aseemingly single row of ovoid, 15-35 pm long stercomatais found in each distal branch.Remarks: Although the species is rarely seen to beintact, its major fragments are so abundant and distinctivethat its description is warranted.Material: H-30.dpoa fragila n. sp.(Pls 9D, 1 iC-D)Diagnosis: Same as that of genus.Description: We possess only one large specimen(PI. 9D), 3 mm long and 2 mm across, that we think isclose to complete; it is characterized by having thecentral tubule surrounded by primary branches.The rather vaguely defined central tubulemeasures 150-190 pm across and may be irregularlysubdivided by constrictions. It is recumbent, as suggestedby the relatively weak development of primarybranches on one surface, presumably the bottom.Few large primary branches extend radially fromthe stalk formed by the central tubule. Their basesare sometimes broadly connected with the centraltubule, but more often constricted and thereforedistinct from it. Perhaps becmse of the constricti~nsthese branches are easily broken off, and thereforethis species is represented almost entirely by isolatedmajor fragments. Primary branches measure100-125 pm across and 125-625 pm in length (frombase to burst of multiple branching).Numerous branches with rounded tips occur atthe distal end of primary branches through dichotomousor multiple branching in closely spaced succession.These branches measure 25-60 pm acrossand up to 375 ym in total length. Some of the mostdistal branches are little more than knoblike processes.In combination, the distal branches give thetotal body a loose, lobular appearance.Normanina Cushman, 1928, emend. hereinNormanina Cushman, 1928, p. 7; Galloway 1933, p.80; Cushman 1948, p. 85; Loeblich & Tappan 1964,p. C192.Diagnosis: Base or center composed of tubules tightlybranching in complex, irregular pattern. Tubulesradiate outward from base and terminate in globularor clublike portion.Type species: Normanina conferta (Norman, 1878).Distribution: Cent.N.Pac.: H-29, H-30, H-36.NW.N.Atl.: Va-9.E.At1.: Bi-DS 78, Me-12-35.Carib.: Ku-1187, Ku-1201.Remarks: Starting with Cushman (1928, p. 7) thegems ,niorw~nim has bee:! variously chxacterizedby different authors. All of them, like Norman, regardedthe organism as a bunch of individuals, andtheir definitions reflect their interpretations of twofeatures, viz., the localization of the oldest part andthe presence or absence of apertures.In his original description of the type species (asHaljphysema conferturn) Norman (1878, p. 279) describedthe tubular portion of the branches as "theirbases", thereby infemng that this is the oldest part.He mentioned apertures as "mouth-opening verylarge", but did not say anything about the location,and his figures do not illustrate this point. Cushman(1928, 1948) interpreted the distal swelling of the

anches as the oldest part, referring to aperturesat the end of the tubular portion. Galloway (1933)seems to have regarded the tubular part as theoldest, placing an aperture at the end of the tubule.Loeblich & Tappan (1964) mentioned "central mass,from which tubular portions radiate9', inferringthat the globular part is the youngest, and they didnot observe apertures in the type specimen.The central or basal part has been interpretedsomewhat differently as "a nearly globular aggregationof pedicels", "mass", "rounded mass", and "centralmass" (resp. Norman, Cushman, Galloway,Loeblich & Tappan, op. cit.). This may come fromthe fact that the central tubule complex is partlycovered with sediment.Although not naming or giving a closer descriptionof the specimens concerned, Brady (1884, p. 276)seems to mention a Normanina sp. from 3950 m be-0.5mmI Itween Juan Fernandez and the South Americancoast as 6. , . an organism . . generally taking theform of little rosettes". The fate of the specimens isunknown (C. 6. Adams, personal communication).On the basis of his studies on a species of Halyphysema,the genus to which Norman ascribed N.conferta, Schepotieff (1912, p. 47) concluded that thisspecies was only a growth form of the widely distributedshallow water species H, tumanowiczii. Thispoint of view has not been discussed by otherauthors, and it is in our opinion entirely wrong.Christiansen (1964, p. 135) reported on some specimens,which he referred to N. conferta, from OsloFiord. Although it is difficult to be sure in the absenceof a detailed redescription of the type materialwhich was not available, we doubt that he dealt withNorman's species; the swellings of the Norwegianspecimens are much too regular in outline, and thebasal part of the tubules were not connected witheach other. The Norwegian specimens are not accessible.Saidova (1968, p. 18; 1969, p. 133 and 134) liststhree names: Normanina fruticosa, N. ultraabyssalica,and N. elongata, which are nomina nuda (1974, personalcommunication).Two types of specimens are abundant in our material.'vVe regard both as new species aiid describethem as N. tylota n. sp. and N. saidovae n. sp.Normanina tylota n. sp.(Figs 5-6; Pls 1 1E-F, 12C-D,F, 19B, 22A-B)Material: H-30.Fig. 5. Normanina tylota n. sp.; type specimen, slightly modifiedbecause of damages.Diagnosis: Body 1.0-1.7 mm in diameter, roughlyspherical. Up to about 60 radiating, 300-600 pm long,fragile, unbranched tubules each bearing a clubshapedto irregular swelling. Swellings internallysubdivided by septae. Between swelling-bearingtubules radiate fewer, easily broken simple tubules,up to at least 1250 pm long.Description: Diameter 1.0-1.7 mm. Specimensroughly spherical although in many cases with asomewhat flattened side, which may be the bottom,or an artifact of preservation. There is a conspicuousdifferentiation into a massive central orbasal part, and an open, peripheral radially organizedpart.The central part consists of a mass of repeatedlydich~tomo~s!y dividing tubules, with sediment fi!!ingin some interstices. Anastomoses between the tubuleswere not seen, but their existence cannot be excludedbecause of the tightness of the mass.More sediment is gradually deposited as the bodygrows, the central part making up 1/5-1/3 of thediameter at a given size. The branching pattern inthis part of the body is, therefore, visible only insmall specimens and occasionally in fragments oflarger specimens. There may be a tendency of thesefrequently branching central tubules to have a slightlylarger diameter than the unbranched peripheraltubules (in a small specimen they measured 31-38 pmagainst 25-31 pm).

Straight or moderately crooked, unbranchedtubules of two types radiating from the central partconstitute a characteristic peripheral area.Most numerous (about 60 per specimen) arecomparatively short tubules each bearing a distalswelling. They are 300-600 pm long measured frombase to beginning of swelling. The diameter is eventhroughout the single tubule, but varies betweentubules of the same individual from 25 to 50 pm,commonly about 35 pm. Branching is rarely seen,and anastomoses were not found. The distal swellingsare set off from the tubules by form and darkercolour. The area of transition from tubule to swellingis rather abrupt (PI. 12C-D), although a more orless flaring part can always be distinguished. There isgreat variation in form and size of the swellings betweenindividuals, and to a lesser degree within thesame individuai. If differences in form and dimensionare expressive of different ages, the development of aswelling can be roughly outlined. The smallest swellingin our material is nicely club-shaped, measuring150 pm in length (direction of tubule axis) and 60 ymin greatest diameter (perpendicular to'tubule axis).When growth proceeds, it is at different rates in differentdirections. There is virtually no increase inlength, whereas the diameter is considerably enlargedin two opposite directions so that a roughlybiradial form is produced (Fig. 6, A-F). The growthis not always equally strong in both directions, moreor less skew swellings being common (Fig. 6, 6-M).In the more regular ones at this stage, the long axis(perpendicular to the tubule axis) often measuresabout 250 ym. During the final growth, up to a size of350-400 pm, irregularities appear in the form of low,rounded knobs at different places on the swelling,seemingly with no relation to the former growthdirections (Fig. 6, N-R). Sometimes two swellings arefused together, an event seemingly occurring only ata relatively early stage of development (Fig. 6,s).The other, less common (10-20 per specimen)tubule type is long, up to at least 1250 ym, and bearsno swelling at the distal end. The diameter is eventhroughout the same tubule, varying from 25 to 30ym between tubules. The distal ends may be open, asif they were large apertures with the same diameteras the lumen of the tubule, or as if the end of thetubule was broken off. Alternatively, they taperslightly with a rounded, closed tip, representingeither a growth stage of this tubule type or theunbroken condition.Colour shades of tan; tubules light, swellings somewhatdarker owing to internal accumulations of ingestedmaterial and stercomata and because ofthicker layer of agglutinated material. Agglutinatedparticles clay and silt. Tubules fragile but neverthelesssomewhat flexible.Tubule and swelling wall in two distinct layers (Pl.22A-B). Inner organic layer

swellings. These thick parts of the swelling wall,seemingly only one per swelling, are quite oftendistinctly set off from the rest of the wall, and mayrepresent accumulation sites for wall material.Cavity of tubules without septae, most often empty,but sometimes more or less filled with stercomata oraccumulated material.Cavity of swellings subdivided by septa which aredouble layered (Pls 12F, 22B). In one preparation,there could be seen numerous small foramina (Pl.19B). Fundamentally, the septa run transversely, asseen in young swellings, but with further unequalgrowth they are turned more or less parallel to theaxis of the tubule, in some cases so much that it looksas if the septa trend toward the stalk. The chambersresulting from septation are somewhat irregular,rounded, and of different size, although they commonlymeasure about 20 pm across. They all share apart of the surface of the swelling.A large mass of stercomata is found in each chamber.They are ovoid and measure 4-9 pm in length.Nuclei may be represented by sparsely distributed,rounded or ovoid, granulated bodies 5-9 pm indiameter, seemingly one in each chamber.Remarks: It seems possible to explain swelling form,pattern of septation, and arrangement and appearanceof the demarcated stercomata masses in termsof plasma behaviour. From the beginning, in small,regular, nonseptate, club-shaped swellings, theplasma string containing accumulated food particlesand stercomata is undivided. The diameter of theswelling cavity increases together with that of theplasma string, and at a certain size, when the diameterreaches about 3 times that of the stalk tubulelumen, the plasma divides, isolating the outermostpart by a transverse septum. The isolated part continuesto live, as indicated by the double nature of theseptum, and presumably is in ccntact with the environmentthrough the swelling wall. As growthproceeds new volumes of plasma become largeenough and subdivide. The septa are still transverselymade with respect to the growth direction of theplasma, but this no longer coincides with the long axisof the tubule which bears the swelling. The plasmastring turns or bulges in the perpendicular directionafter having separated the first or a few chambers,thus producing skewed and biradially swelling forms.The final irregular stage may be reached either byfurther growth and dividing of the original plasmastring, by growth and dividing of the isolated plasmalumps in the chambers, or by a combination of thetwo. What makes the growth direction of the plasmachange is difficult to say; it may be a question of theshortest distance to the surrounding water, or thelocation of the accumulation sites, or it may be an intrinsicfactor related fo the segregated plasma lumps.N. tylota seems to be very near to the type speciesof the genus, N. conferta, which is known only fromthe type locality in the North Atlantic (Norman1878). However, the descriptions as given byNorman and later authors are very general anddevoted only to the external appearance. Moredetailed information on the holotype of N. confertawas provided by C. G. Adams (personal communication).Important differences between N. tylota and N.conferta are the presence of two types of tubules andlarge variation in swelling form in the former, whilethe latter has only one tubule type and comparativelyregular swelling form. Other differences are foundin the length, diameter, and number of tubules, andin the grain size of agglutinated particles.Material: H-30.Normanina saidovae n. sp.'(Fig. 7; Pls 12E, 23A-C)Diagnosis: Body 0.7-1.6 mm in diameter, irregularlyrounded, somewhat oblong. Numerous tubules, oftenmore than 100, flexible, radiating, up to 600 ym long,repeatedly dichotomously divided, each bearing aclub-shaped, nonseptate swelling.Description: Body 0.7-1.6 mm in diameter, irregularlyrounded, often somewhat oblong. Conspicuouslydifferentiated into a central massive part and aperipheral, radially organized part.The central part consists of a single broad tubuleor bulb. The diameter is about 100 pm and the lengthup tc about ! mm. In EOE~ czses, detzils of the centralpart are obscured by the numerous radiatingtubules. In sections, the surface shows 3-4 deep constrictions(Pl. 23A) which together with radiatingtubules branching off from the intermediate nodesare reminiscent of a short piece of backbone. In onecase the tubule was seen to branch.The peripheral part consists of radial, straight orsomewhat crooked, usually dichotomously dividedtubules bearing club-shaped swellings at the end of thebranches (PI. 12E). At the base, the radiating tubules1. This honours Ch. M. Saidova for her comprehensive deep-seaforaminifer research.

Remarks: N. saidovae is characterized by the easilyvisible, repeated branching which results in a largernumber of radiating tubules, and by the regularity ofthe club-shaped swellings. In N. tylota and N. confertabranching of the radiating tubules is very rare, thenumber of swelling-bearing tubules is only about halfof that of N. saidovae, and the swellings are of irregularform. From N. tylota, N. saidovae further differsin the organization of the central part and the lack ofseptation in swellings.Lana n. gen.I0.5rnrnFig. 7. Normanina saidovae n. sp.IDiagnosis: Loose clump of branching, anastomosingtubules; no focal point for symmetry or growth pattern.Interstices not filled with sediment. Tubulesstraight-sided (not constricted), nonseptate.Type species: Lana neglecta n. sp.originate from the central tubule in a number of 6-10per node. Secondary and tertiary branching is verycommon, generally being well-spaced. Quaternarybranching seems rare. The exact pattern ofbranching is difficult to see without destroying thespecimens; only the outer one or two generations ofbranching are readily visible. The total length of thebranched tubules, from central tubule to tip of swelling,varies greatly and may be up to 600 pm. Thediameter is 25-40 pm, and it is not markedly influencedby the branching. The diameter of the swellings(measured as largest diameter) is 75-100 ym.Because they taper gradually into the tubule, theirlength cannot be given. Fusion of two swellings israrely seen.Colour tan, but stercomata usually visible as greycore. Agglutinated particles clay and silt. Tubulesflexible.Ix centra! %bale wall, inner organic hyer 0.5-1.5ym thick and outer agglutinated layer 10-20 pm thick.The cavity is incompletely subdivided by the constrictionsin a short series of chambers. Both layers in thewall take part in the constrictions (PI. 23A) whichaccordingly is not comparable to the septa found inother species. The cavity is nearly empty, only scatteredstercomata being found.In walls of tubules and swellings, inner organiclayer

has not been found. We reject their inclusion here cases the tubules seem to be flattened and empty.because too few details are known, and, more im- Some sectors of a clump may appear grey, whileportant, the dimensions of body, tubules, and agglu- others verge toward tan; in such cases grey tubulestinated particles are much larger than in Komo- may be slightly thinner. Sides of tubules parallel, notkiacea.constricted, although sometimes shriveled. Tips oftubules rounded. Agglutinated particles clay.Tubule wall with thin inner organic layer 4.5 ymLana neglecta n. sp.thick, and thicker agglutinated layer 1-5 ym thick.(Fig. 8; Pls 13A-B, 14D, 26B)Outer layer in some parts of a specimen absent, possiblyas the result of handling.Material: H-30.Cavity of tubules nonseptate, with stercomatavarying in quantity from sparse in one area toDiagnosis: Tubules 20-25 ym in diameter. Distance abundant in another. stercomata ovoid, 4-10 ym inbetween branchings large in comparison with dim- diameter. Occasionally the tubule cavity is filled witheter of tubules (ratio commonly >6). ~ olo~r basically ingested material not bet?) formed as stercomata.grey.Blurred bodies resembling nuclei, spherical orovoid, measuring 2-4 ym in diameter, were foundDescription: Body is a loose, disorganized tangle of sparsely in all specimens sectioned.branching tubules forming a clump, up to 5 mmacross, with rather ill-defined surface (PI. 13A-B).Density of clumping somewhat greater toward cen-Lana reticulata n. sp.ter. Branching irregular, ranging from acute, (Pls 14A-C, 19C)through right-angled, to obtuse; at juncture, twobranches may diverge from direction of third, or Material: H-30.one may extend straight on from another, with thethird coming off at angle (Pi. 14D). Branching gener- Diagnosis: Tubules 30-50 ym in diameter. Distanceally dichotomous; successive branchings may be very between branchings relatively short in comparisonclose together (3 tubule diameters) but are most with diameter of tubules (ratio commonly about 4).frequently a long distance apart (commonly 6 tubule Colour tan.diameters). Between branchings, tubules nearlystraight, crooked, or strongly curved.Description: We have not been able to discern whatTubules stiff, but flexible, all of same basic diam- a complete individual looks like. Most of our specieter,20-25 ym. Larger diameters occur, but in these mens are spheroidal, 2-4 mm in diameter, but actualsurficial layers may have fragmented off all of them.Distance between branchings most frequently 3-5tubule diameters, or less. Branching dichotomous,although adjacent branchings may be so close togetheras to look multiple. Tubules tend to diverge atequal 60" angles, and interbranching distance rela--1 mmFig.1868. Lana neglecta n. sp.

tively constant, so that the openings of the meshworkare fairly symmetrical in shape and constant in size(Pl. 14A-C).Most individuals evenly tan, reflecting generalpaucity of stercomata within the tubule cavities.Tubules flexible, but stiff. Agglutinated particles clay.Inner organic layer of tubule wall 0.5-1 pm thick,outer agglutinated layer 3-5 Fm, occasionally up to 10pm thick.Cavity of tubules nonseptate, with only small butconstantly occurring quantities of ingested materialand spherical to ovoid stercomata which are up to 8pm in diameter.Nucleilike bodies spherical or faintly ovoid, 6-10pm in diameter, with granulated interior.Remarks: Differences separating L. reticulata fromk. neglecta are that its tubule diameter is larger, itsbranching is more regular, and the openings of themeshwork smaller, more constant in size, and moresymmetrical in shape.Our collections contain several undescribed specieswhose morphology approximates that of L. reticulata.Station H-30 contains at least three suchspecies. Their presence in the same sample in sufficientabundance allows recognition of the discrete,albeit subtle morphological gaps that separate them.The criteria used are tubule diameter, distancebetween branchings, regularity of branching, orientationof tubules, integrity of the total mass, andcolour. However, the differences can be so subtlethat direct comparison is advisable wherever possible.BAGULELLDAE n. fam.Diagnosis: Body variously shaped: straight shaft,treelike and sparsely branched, or dense clump.General fine morphology dominated by beadlikeappearance which is caused by frequency of extremelyshort branches, often constricted at base,and by frequent constriction of main tubules at regular,short intervals.Type genus: Baculella n. gen.Remarks: The genera are presented in an orderreflecting increasing complexity in primary tubulemorphology.Treelike, sparsely branched species are known tous, but not described in this report.Baculella n. gen.Diagnosis: Body a single, sometimes dichotomouslybranched shaft, formed by a single nonseptate, butsometimes constricted tubule from which ariseshort side branches which may be simp1ebeadlike lobes.Type species: Baculella globofera n. sp.Distribution: NE. N.Pac.: Vi-6104, Vi-6118.Cent.N.Pac.: H-30, H-37.E.eq.Pac.: Gal-716.E.Atl.: Bi-DS78, Me-12-35.Carib.: Ku-l178A, Ku-1187.W. eq.Atl.: Gal-52. ,W. Indian: AB-382C.Remarks: The diagnosis covers a fairly large numberof species, and it may be possible that the genusshould in future be subdivided into at least two subgenera,one represented by B. globofera n. sp. andone by B. hirsuta n. sp.In the diagnosis we have included the possibilitythat the shaft of the body may be dichotomouslybranched; this character is not seen in the two speciesdescribed here, but it is found in other, undescribedspecies.Material: H-30.Baculella globofera n. sp.(PIS 15A-B, 16A, 19D, 24A-B)Diagnosis: Body straight or slightly bent, up to about3 mm long and a little less than 0.5 mm in total diameter.Side branches relatively sparse and simplybranching, with no sediment in interstices so thats+,ll .+- -1 L..l-\ -LCUK (C~ULILII LLIUUICJ is easily visibie, even in iargeindividuals. Colour tan.Description: Body up to 2.8 mm long, always ratherstraight, and 250-430 pm in total diameter. The diameteris nearly constant through the whn!e length, 1-though some individuals are a little thicker in themiddle part, others at one end. The shaft or centraltubule is 60-125 pm in diameter, the lowest numbersrepresenting modest constrictions at intervals of90- 120 pm.About half-way between the constrictions 2-5primary branches arise all around the centraltubule, more rarely concentrated along one side.

They measure up to 200 pm in length (generallyabout 125 pm) and 40-60 pm in diameter. Secondaryand tertiary branching is frequent, with distancebetween branchings usually no more than one diameter,rarely as much as three diameters. Terminalbranchlets no longer than wide, broadly rounded,and broadly based rather than constricted. Primarybranches not packed together; perhaps as a result,orientation of secondary branches follows no preferreddirection.In young specimens the stalk tends to zigzag, withbeadlike or short tubular side branches arising ateach salient; this pattern is obscured with age.Colour tan, with rounded masses of stercomataseen as small, dark spots inside the branches. Flexibleand elastic. Agglutinated particles clay.In central tubule the inner organic layer is about0.5 pm thick, sometimes more, and the outer agglutinatedlayer is 10-20 pm thick. In the branches thetwo layers are

Fig. 9. Baarlella hirsuta n. sp.; type specimen.in diameter, than the free end and has fewer smallprotruding points. It is not possible to discern furtherdetails because of the tight spacing of the beadedtubules and the filling of interstices with sediment.Sections reveal that the organizational pattern is apossibly meandering central tubule with branchesarising all around it for its full length.Spacing of the primary branches is exceedinglytight, only 10-25 pm, a fact that obscures the discreteexistence of the central tubule. The diameter of thecentral tubule cavity is 15-25 pm, and the thickness ofthe organic layer 5-10 pm. An agglutinated layercannot be distinguished from the sediment in the interstices.The cavity is nonseptate and seems unconstricted.The branches are round or somewhat flattened incross section, 35-60 pm in diameter and 100-125 pm inlength. They look constricted where they leave thecentral tubule, but not markedly so. Each branch isprovided distally with one or often two blunt orsharp projections (Pl. 24C). These projections whicharise withwt cc=nstrictions are of widely varyinglength, from barely visible up to about 120 pm, andgenerally 5-10 pm in diameter with circular crosssection. Were they to be interpreted as a set of secondarybranches, the branching pattern would haveto be characterized as having a very marked, stepwisedecrease in diameter.Inner organic layer 3-5 pm thick in branches and2-3 pm in distal projections. As in the case of the centraltubule, a special agglutinated layer cannot bediscerned from the interstitial sediment filling. Theextraordinary thickness of the organic layer seemsto result from a secretion of layer upon layer offibrous or membranous matter on the outside of thetubules (Pl. 1%). Here and there a shrinking hasoccurred, causing an inner lamina, (0.5 pm thick, toseparate from the rest of the wall. The thickness ofthe organic wall decreases toward the sedimentcoveredend, in some places down to about 1 pm.Cavity of branches and projections nonseptate andnot constricted.Cavity of central tubule and branches totally butloosely filled with ovoid, up to 9-pm-long stercomata,here and there with patches of loose, ingested material(PI. 25.4). There is a tendency for this materialto be more common in the sediment-covered end.Stercomata are sometimes found in the projectionsalso, but more often these are empty.Bodies resembling nuclei are found scattered;they are spherical or ovoid with faintly granulatedinterior and measure 5-7 pm in diameter.This is one of the few species in which the plasma isrelatively easily seen.The fibers in the sediment sheath arise at the surfaceof the body. The thickest are about 10 pm indiameter, the thinnest about i pm. Thickness seemsto be achieved through depositing many thin fibers ina bundle. Dichotomous branching is frequent butirregular, so that parts are not equally thicks; it goeson through splitting of the bundle, not through divisionof the single fiber. NQ rnastcxmses were seer,.The fibers appear to be of the same material as theorganic layer of the wall. Sediment particles seem tostick to the fibers and are sometimes found inside thefiber bundle.Remarks: B. hirsuta differs from B. globofera in alarge number of characters. The most obvious, externallyvisible ones are: general body form, pres-

ence of the fibers and the sediment cap, sediment fillingof interstices between branches, and presence ofthe small conical projections.Because the primary tubule is of a remarkablysmall diameter in B. hirsuta and shows marked constrictionsin B. globofera, the organization of theformer might be interpreted in another way thanabove. What is here called central tubule andbranches could, in fact, be a case of an exceedinglystrong and tight constriction of a central tubule, withthe distal projections representing the first and onlygeneration of branches. The branching patternwould then be characterized by the marked decreasein diameter following the first step ofbranching. Only investigation of a number of otherspecies will show if this interpretation is correct.It appears from the description that the two endsof an individual are dzferent, with gradual changesalong the central tubule. These changes are seen inthe sediment covering and fiber distribution, bodydiameter (without sheath of sediment), thickness ofthe organic layer in the test wall, and ratio of ingested,loose material to quantity of stercomata. Ourinterpretation is that the thick, sediment-coveredend is the younger, and that the sheath is graduallycast or eroded away from the older end. Plasma andsupposed nuclei are most often found in the youngerend, but they are seen throughout the whole individual,and although we think that the activity is greatestin the young end, it should be taken into considerationthat the more tightly packed stercomata in theold end make detection of other structures difficult.Edgertonia n. gen.'Diagnosis: Body formed as dense, rounded clump ofbranchmg tubules with numerous extremely shortside branches and beads.Type species: Edgertonia tolerans n. sp.Distribution: NE.N.Pac.: Vi-6 109.SE.N.Pac.: Vi-4281.Cent.N.Pac.: H-29, M-30, H-36.E.eq.Pac.: Gal-716.Carib.: Ku-1178A, Ku-1187, Ku-1189.W. Atl.: Eastw-20087.E.At1.: Bi-DS78.1. Ths genus is named in honour of Carol C. Edgerton, whosecareful sorting led to the first appreciation of the superfamily andto our finest collection of material.Material: H-30.Edgertonia tolerans n. sp.(PIS 9E, 16B-F, 26C)Diagnosis: Body up to 4.5 mm in diameter. Tubulesanastomosing. Beads usually spaced so that outlineand course of single tubules can be traced.Interstices without sediment. Cavity of tubulesnonseptate.Description: Round or slightly distorted spheroidcomposed of tangle of closely branching, anastomosingtubules. Undamaged individuals up to 4.5mm in diameter, and we have fragments possiblyoriginating from even larger specimens. No sedimentin interstices, giving appearance of a bathsponge. Meshwork of tubules clearly open, butmoderately dense so that visibility to the interior isblocked after only 2-3 layers. Surface distinct, notfrayed, except where specimen appears damaged. Novisible pattern of orientation of tubules. No free tubulesextending from surface.Tubules 25-40 pm in diameter, usually dichotomouslybranchng, but sometimes more. Short sidebranches protrude from surface at short, irregularintervals (from a few to 125 pm) (Pl. 16B, D-F).These side branches frequently simple spheroids("beads"), 35-50 pm in diameter, with constrictedbases, or short branches, not much longer thandiameter, unconstricted at base and terminating incluster of beads. Beads and side branches may comeoff singly or multiply.Colour tan. Consistency markedly spongelike, withtubules resisting tearing off. Agglutinated particlesclay.Tubule wall with inner organic layer (0.5 pm thick.Outer agglutinated layer 2-12 ym thick, generallyabout 5 pm.Cavity of tubuies nonseptate, narrowing at beadconstrictions. Some tubules empty, others well filledwith stercomata and ingested material. Stercomataovoid, up to 8 pm long.Nuclei fairly numerous, spherical, and with granulot-rl;ntpr;nr 1-2 ;,rn in Aiam~t~ro~q~r~lly about 5crrbu IIILUIIVI, _I u pI1l 1. UIU.l..,.r., ---A-m.Edgertonia argillispberula n. sp.(Fig. 10; Pls 9F, I7A-B, 18A-B, 25B, 26D)Material: H-30.

Diagnosis: Body up to 3.5 mm in diameter. Tubules The distal ends are broadly open, possibly as a resulttightly covered with beads so that outline and course of breakage, or gradually tapering toward a bluntof single tubule is difficult to trace. Interstices filled tip. Although in some specimens these tubules springwith sediment so that appearance is much like that of from all parts of the surface, there is in most cases aa mud ball. Cavities of body tubules infrequently rather clear restriction to an equatorial belt.septate. Long, unconstricted tubules may extend Colour greyish tan. Rather soft, not flexible. Agfreelyfrom body surface.glutinated particles clay.Inner organic layer of tubule wall

from the surface of many of the individuals may bean age phenomenon since it is most common in largespecimens, or simply an artifact because of theirfragility.The most obvious differences between E. toleransand E. argillispherula are the lack of sediment in theinterstices, the spongy consistency, and the spacedbeading of the former, as opposed to the sedimentfilledinterstices, the unflexibility, and the tightbeading in the latter. Other differences of the latterare found in the frequent presence of long free tubules,the presence of septa in the tubule cavities, andin the the degree of adherence of the tubule latticework.RII. SUMMARYThe aim of this paper is to introduce and establish ahitherto overlooked group of benthic deep-sea animals,the "komoki", belonging to the Foraminiferida.Because the group has turned out to be surprisinglyrich in species, this first treatment is restricted to ageneral characterization, the erection of a newsuperfamily, the Komokiacea, and the descriptionof some lower taxa which typify the range of morphologiesfound in the material at disposal.Komoki have only in recent years been securedin large numbers. The reason is that they are so fragileand small that their retention generally demandsgentle washing through a fine-mesh screen. In thepresent work, they have been studied by use of dissectionmicroscope, scanning electron microscope,and paraffin embedding followed by sectioning andgeneral staining methods. The terminology used isthe same as for other Foraminifera. A list of elevenpoints, to which attention should always be paid whendescribing new species of this group, is set up (p. 171).The core of the material originates from the areaunder the central North Pacific gyre. The effortswere concentrated on this material because one canthen see better the range of morphologies that livetogether than if the species are drawn from differentplaces. Moreover, the way in which it was collected isknown in technical detail, and we have a comprehensiveknowiedge about the comiiiuni:ji from which itcomes.After having acknowledged the reality of thegroup, it soon turned out that numerous, generallysmall, samples exist in the materials from many expeditionscarried out in different parts of the worldocean. Thus the experience laid down here, especiallyconcerning the distribution of the genera, could bedrawn from more than 100 samples.Komoki generally measure 1-5 mm in Iargestdimension. The basic element in their morphology isa fine tubule of even diameter, containing the plasmaand a large quantity of faecal pellets (stercomata).External appearance is above all influenced by tubebranching pattern and presence of sediment betweenthe tubules in some species. Two types of branchingseem to be of constant occurrence. In the first thetubes are cylindrical, rarely constricted, and withlarge distances between branching points. In thesecond type is found a rich branching, resulting in alarge number of close-set, very short side branches,and the main tubules are frequently constricted.The tube wall consists of two layers, sharply demarcatedfrom each other. The interior layer isgenerally

samples at depths of 1000-7000 m. The conclusion isthat komoki is essentially a deep-sea group, with itsshallowest distribution somewhere on the continentalslope, perhaps correlated with abrupt changes intemperature and sediment structure. Two genera(Septuma and Lana) are reported from bathyaldepths, all seven genera have their main distributionin the abyssal, and four genera (Ipoa, Lana, Bantlella,and Edgertonia) extend into true hadal depths.The greatest relative abundances of komoki havebeen found in the oligotrophic abyss and in hadaltrenches, where their volume exceeds that of all themetazoans combined and equals that of the remainsof other Foraminifera. As a first approximation todocument the diversity of komoki and their distributionin the bottom, the results of analyzing 500 cmZ ofa box core are given: about 2000 specimens werefound that could be sorted into 56 species, of which32 belong to two genera. More than 80 % of the specimenswere in the top 1 cm layer, whereas only2.5 % were below 2 cm. Some species show greatestabundance in the 1-2 cm layer, and five species werenever found in the 0-1 cm layer, indicating that atleast some species live well beneath the surface. Oneconclusion of this is that much caution should be usedin deducing life style from morphology in komoki.Within the order Foraminiferida, komoki areplaced in the suborder Textulariina as a new superfamily,Komokiacea. Two new families, Komokiidaeand Baculellidae are erected, respectively with fiveand two genera. A key is prepared for families andgenera. In Komokiidae all species and genera butone are new: Komokia n. gen. (with K. multiramosa n.sp.), Septuma n. gen. (with S. ocotillo n. sp.), Ipoa n.gen. (with I.fragila n. sp.), Normnina Cushman, 1928(with N. tylota n. sp. and N. saidovae n. sp.), and Lanan. gen. (with L. neglecta n. sp. and L. retintlata n. sp.).In Baculellidae all species and both genera are new:Bantlella n. gen. (with B. globokra n. sp. and B. hirmtan. sp.) and Edgertonia n. gen. (with E. tolerans n. sp.and E. argillkpherula n. sp.).PX. REFERENCESAllen, J.A. & H.L. Sanders, 1966: Adaptations to abyssal life asshown by the bivalve Abraprofundorum (Smith). - Deep Sea Res.13: 1175-1184.Brady, H.B., 1884: Report on the Foraminifera. - Rep. scient,Results explor. Voyage Challenger 9: 1-814.ChPistiansen, B.O., 1964: Normania conzrtum from the OsloFiord in Norway . - Contr. Cushman Fdn foramin. Res., 15:135-137.Cohen, A.L., D.P. Marlow & G.E. Garner, 1968: A rapid criticalpoint method using fluor carbons ("Freon") as intermediateand transition fluids. - J. Mlcrosc. 3: 589606.Cushman, J.A., 1928: Additional genera of the Foraminifera. -Contr. Cushman Fdn foramin. Res. 4: 1-8.- 1948: Foraminifera. Their Classification and economic Use. -Harvard University Press. Pp. 605.Fenchel, T., 1972: Aspects of decomposer food chains in marinebenthos. - Verhandl. dt. zool. Ges., 65. Jahr. vers.: 14-22.Galloway, J.J.: 1933: A manual of Foraminifera. - PrincipiaPress. Pp. 483.Haeckel, E., 1889: Report on the Deep-Sea Keratosa. - Rep.scient. Results explor. Voyage Challenger 32: 1-32.Hessler, R.R., 1974: The structure of deep benthic communitiesfrom central oceanic waters. - Pp. 79-93 in: The biology of theoiealic Pacific. Roc. 33rd Ann. Eioi. Coii. Oregon Srare iiniversityPress.Hessler, R.R. & P.A. Jumars, 1974: Abyssal community analysisfrom replicate box cores in the central North Pacific. - DeepSea Res. 21: 185-209.Honigberg, B.M., W. Balamuth, E.C. Bovee, J.O. Corliss, M.Gojdics, R.P. Hall, R.R. Kudo, N.D. Levine, A.R. Loeblich, J.Wieser & D.H. Wenrich, 1964: A revised classification of thephylum Protozoa. - J. Protozool. 11: 7-20.Ivanov, A.V., 1963: Pogonophora. - Academic Press. 4.479.Johansson, K.E., 1937: Ober Lamellisabella zachsi und ihre systematischeStellung. - Zool. Am. 117: 23-26.Jumars, P.A., 1975: Methods for measurement of communitystructure in deep-sea macrobenthos. - Mar. Biol. 30.245-252.Lemche, H., 1957: A new living deep-sea mollusc of the cambro-devonianclass Monoplacophora. - Nature, Lond. 179:413-416.Lemche, H. & K.G. Wingstrand, 1959: The anatomy of Neopilinagalatheae Lemche, 1957 (Mollusca Tryblidiacea). - GalatheaRep. 3: 7-7 1.Lengsfeld, A.M., 1969: Nahrungsaufnahme und Verdauung beider Foraminifere Allogromia laticollaris. - Helgoliinder wiss.Meeresunters. 19 385-400.Loeblich, A.R. & H. Tappan, 1964. Treatise on InvertebratePaleontology. Part C, Protista. - University of Kansas Press. Pp.900.Luna, L.G., 1968: Manual of histological staining methods at theArmed Forces Institute of Pathology. - McGraw-Hill Book Co.Pp. 258.Marszalek, D.S., 1969: Observations on Iridia diaphuna, a marineforaminifer. - J. Protozooi. 16: 599-61 l.Nemanic, M.K., 1972: Critical point drying, cryofracture andserial sectioning. - Pp. 298-304 in: Scanning Electron Microscopy.Proc. Fifth Ann. Scanning Eiectron Microscopic Symp.Newell, R., 1965: The role of detritus in the nutrition of twomarine deposit feeders, the prosobranch Hydrobia ulvae and thebivalve Macoma balthica. - Proc. zool. Soc. Lond. 144.2545.Norman, A.M., 1878: On the genus Haliphysema, with descriptionof several forms apparently allied to it. - Ann. Mag. nat.Hist. (5) 1: 265-284.NGrrevang, A,, 1970: The position of the Pogonophora in thephylogenetic system. - Z. zool. Syst. & Evolutionsforsch. 8:161-172.

Saidova, Kh.M., 1968: Foraminifera. - Pp. 17-26 in: The PacificOcean VII: Biology of the Pacific Ocean. Book 11. Part 1: TheDeep-Sea Bottom Fauna. Moscow. (In Russian; translated 1970by U.S. Naval Oceanographic Office, Washington, D.C.).- 1969: The distribution and ecology of the recent benthonicForaminifera in the Pacific. - Pp. 120-201 in:The Pacific Ocean.The microflora and microfauna in recent sediments of thePacific Ocean. Moscov. (In Russian).Sanders, H.L., R.R. Hessler & G.R. Hampson, 1965: An introductionto the study of deep-sea benthic faunal assemblagesalong the Gay Head-Bermuda transect. - Deep Sea Res. 12:845-867.Schepotieff, A., 1912: Untersuchungen iiber niedere OrganismenI. Die Gastraaden (Haliphysema und Gastrophysema). -2001. Jb. Anat. Ontog. Tiere 32: 43-76.Tendal, O.S., 1972: A monograph of the Xenophyophoria(Rhizopodea, Protozoa). - Galathea Rep. 12: 7-99.Williams, G.E., 1974: New technique to facilitate hand-pickingmacrobenthos. -Trans. Am. microsc. Soc. 93: 220-226.