profiles to be plotted on lIlaps—a technique particu- larly helpful in reconnaissance surveys (see, for An Operational example, Herron, in press). Processing System for The production of charts merits further discussion because of the potential application of this tech- Natural History Specimens nique to other investigations. For instance, final ad- justed track charts could be produced on board ship on a- day to day basis. The programs incorporate B. J. LANDRUM subprogFams which provide scaling, grid drawing, 2 Smithsonian Oceanographic Sorting Center and annotating for a Mercator Projection of vari- Smithsonian Institution able scale. Functions are employed which convert latitude and longitude to the proper x-y coordinates and for plotting within a specified area. These programs K. G. SANDVED utilize tables (within the program) which are effec- tively tables of meridional parts and allow plotting Office of Antarctic Programs of all areas of the world from 80°N. to 80°S. The National Science Foundation size of a specific plot is limited only by the scale chosen and the physical limitations of the plotter. In this fast-moving world where advances in It is the intention of the Lainont-Doherty investi- technology often seem to outpace the progress in gators eventually to do all basic and science, there are some areas where the phase dis- display on board Eltanin within a few hours or a placement is particularly evident. A good example day of the actual data collecting. There are many is museums, which, although faced with formidable advantages to this approach: Errors in data collec- problems in inventorying and cataloging specimens tion will be caught early and corrected, more intelli- collected over decades and even centuries, only re- gent decisions regarding the day-to-day and cruise- cently have begun to apply to the task to-cruise schedule and track of the ship can be made, of inventorying and retrieving their ever-increasing and time in the laboratory can be devoted entirely holdings. As a result, large and significant collections to the detailed analysis of the data collected. Experi- exist about which little is known. Thus, in an age ence indicates that the machine time required for the when even railroad operations have become com- reduction and display of the navigation and of the puterized, a systematist studying natural-history geophysical data is less than one hour per day of specimens may still have to go from museum to data collection. museum to view endless rows of labelled containers merely to see if there might be specimens available that would be useful in his research. Depending upon 2 The inappropriateness of the Mercator Projection for the his specific needs, he may conclude that it is easier, southern portion of antarctic waters is readily apparent. and possibly more profitable, to undertake a field- There is no reason why conic or polar projections could collecting excursion of his own, or to participate in not he employed, however; in fact, a Lambert conformal conic projection has already been programmed but has a large-scale, multipurpose program. not yet been widely utilized. In the Lamont-Doherty If he happens to join a marine expedition, our plotting programs, subprograms have been used exten- systematist, who may be concerned with only one sively, and plotting at other map projections can he taxonomic group of animals or plants, will find that realized with no basic changes in programming logic. the use of trawls and similar collecting devices often produces assemblages that far exceed his immediate References needs and interests, while time and shipboard facili- ties may limit efforts to preserve and record any but Hayes, D. E., M. Talwani, and J . L. Worzel. 1967. Sat- the needed materials. The collector may elect to ellite Navigation in Relation to Surface Ship Gravity Measurements. Paper presented at the 37th International return everything to the sea that is not of direct Meeting of the Society of Exploration Geophysicists, interest to him, or he may remove wanted specimens, Oklahoma City, October-November 1967. briefly label the remainder and ship it to his home Herron, E. In press. Crustal plates and sea floor spread- institution, where it may rest with its presence un- ing in the southeastern Pacific. Antarctic Research Series. known to most of the scientific community. Thus, the Talwani, M. 1969. A System for the Reduc- tion, Storage and Display of Underway Data Acquired opportunity to utilize the total assemblage by chan- at Sea. Lamont-Doherty Geological Observatory of Co- neling "unwanted" specimens into the hands of other lumbia University. Technical Report No. 1, CU-1-69 specialists is lost. N000 14-67--A--0 108-- 0004. To assure optimum efficiency, a multipurpose, Talwani, M., J . Dorman, J. L. Worzel, and G. M. Bryan. 1966. Navigation at sea by satellite. Journal of Geophy- large-scale, and continuous expeditionary effort re- sical Research, 71(24): 5891-5902. quires a focal point for coordinating all phases of

278 ANTARCTIC JOURNAL work involved in the care, preservation, recording, made to channel representative specimens into the and distribution of the natural-history collections. collections of the U.S. National Museum (Sandved, Without prompt identification of the specimens, the 1965). In the meantime, large collections had been worthwhileness of the effort is dissipated. Conse- brought back. Some of them were stored or forgotten quently, it is vital to place the unstudied materials for lack of interest, funds, or time to study them. quickly in the hands of specialists. To expedite this Some were examined, then disposed of at the con- process, three ingredients are required: (1) a cen- clusion of the research project. Others were left tral receiving and sorting repository with technical behind in the no-mans- land of university storage competence to match the spectrum of materials, (2) cabinets when the collector moved on to new posts a data-handling facility capable of storing and re- and different challenges. In these and many other trieving essential to the study of the cases, the collections remained inaccessible to other collections, and (3) a body of senior scientists corn- scientists who wanted to study them, quite often petent to evaluate and recomniend the most appro- because their existence was known only to a few priate systematist for each group of specimens. individuals. The problems of data and specimen handling were The bulk of the collections brought back from recognized early in the U.S. Antarctic Research Pro- Antarctica are still distributed widely to individuals gram (USARP), perhaps because the program is a and institutions for identification and study. The prime example of the multipurpose, large-scale significant differences today are that preliminary sort- expeditionary effort requiring detailed coordination ing has been made before distribution, and after the for optinmumu efficiency. The Smithsonian Institution, specimens leave the distribution center, efforts are very much aware of the same problems, decided to made to ascertain any further changes in their loca- establish its Smithsonian Oceanographic Sorting tion and to determine the stage of processing. Center (SOSC) at about the same time (Wallen, During April and May of 1963, two meetings were 1963). held between representatives of the Smithsonian In- The U.S. Antarctic Research Program is a national stitution and the Office of Antarctic Programs of program in which academic institutions across the NSF for the purpose of discussing an arrangement land participate along with government agencies and, whereby the Smithsonian Institution would assume a few commercial organizations, it is financed en- the responsibility for sorting, distributing, and main- tirely with government funds and managed by the taining records on natural-history specimens col- National Science Foundation (NSF), which, through lected under USARP. These exploratory meetings its Office of Antarctic Programs, also provides for led to further discussions and on November 1, 19631 the required support of the scientists that it funds NSF made a grant to SOSC to begin processing of (Sandved, 1965). the marine biological collections from Eltanin and In the field, USARP is supported logistically other vessels operating in antarctic waters. It was largely by the efficient and capable U.S. Naval Sup- soon followed by another grant for the recording of port Force, Antarctica; however, in addition to spe- data on these and other antarctic specimen collec- cialized logistics, there are numerous other support tions. Both projects are now carried out under con- services that must be provided by the National Sci- tract with NSF.1 ence Foundation. Bsides such obvious, major re- NSF supports similar antarctic records centers at quirements as the operation of the NSF research other institutions. For example, a core library and ships Eltanin and Hero, the scientists, for example, distribution center for antarctic bottom-sediment need special clothing for use in the field; advice cores exists at Florida State University, and the and assistance both in the field and at the staging Armys Cold Regions Research and Engineering point in Christchurch, New Zealand; laboratory and Laboratory at Hanover, N.H. assumes a similar role transportation equipment and facilities; and aid for the continental ice cores. With a contract signed while working up their data in the United States. last year with the Smithsonian Institution for the In this brief introduction, we shall focus on one of processing of rock samples at SOSC, the circle ap- these latter services, available to scientists before they pears to have been completed, and present investiga- depart for Antarctica and after they return, and to tors and future generations are assured of access to, those who may never make the trip at all—informa- and preservation of, the antarctic collections obtained tion on natural-history specimens and collections. at the expense of the U.S. Government. The centers enhance the value of the collections by making repre- As a result of the nationwide participation of sci- entists in the antarctic program, the USARP and, particularly, pre-USARP natural-history specimen collections are scattered throughout the United Annual reports on these operations are given in the Sep- tember-October issues of time Antarctic Journal of the States and abroad. Not until 1963 were serious efforts United States.

November—December 1969 279 and new members are generally solicited on the basis of committee recommendations. It was realized from the outset that an automatic data processing system would be a prime requirement at SOSC in order to process the cumulative antarctic records efficiently, and the development of such a system was pressed forward concurrently with the build-up and maintenance of manual files. The plans for the system included one important auxiliary ob- jective: although the system was to be designed for a specific need, it should not become an "antarctic" system in the sense that it could not be used for col- lections from any part of the world. It is hoped, therefore, that the system described in this article will be of interest to anyone trying to cope with the records on large collections of natural-history speci- mens.

The Records System

There were a number of requirements for the type of data processing system that was developed at SOSC. First, it was to be basically an inventory con- trol system, distinguished by the requirement to record Smithsonian Institution Photo information according to the arbitrary and dynamic taxonomic hierarchies that are used by scientists to Figure 1. View of stack area at SOSC housing marine collections. classify natural-history specimens. Second, it should be responsive td specific questions as well as to a sys- investi- sentative samples available to any bona fide tematic readout of the entire data bank. Thus, one gator, and reduce duplication of work in the collect- user might wish to know if specimens of a given taxon ing as well as in the study of specimens. have been collected below a certain depth in a par- The Smithsonian Oceanographic Sorting Center ticular area, while another might require listings of (SOSC) now serves as one of the major sorting, all recorded antarctic invertebrates. Third, the cap- distributing, and recording centers for USARP ture of information should not impose additional natural-history collections. One of SOSCs advant- workloads on the Sorting Center technicians during ages from a government (and university) adminis- the processing of specimens. And fourth, the system, trators point of view is its impartiality: i.e., it has while designed to fit into the normal operations of no research objective or other requirements for the Sorting Center, should be sufficiently flexible to a priori demands on the collections. In fact, its sys- permit inclusion of records from any other source tern of distributing specimens to specialists with the with minimal changes in either procedures or data aid of advisory groups was developed to assure fair manipulation. treatment of interested scientists and optimum utili- zation of the collections. At present, eight advisory committees, each usually with five members, con- The Unit Record sider requests for specimens and provide the SOSC with recommendations for distribution of the speci- The rationale for the system development rests in mens. The eight committees are designated by tax- part on the fact that the vast majority of natural- onomic groups or specialties as follows: Algae, Ver- history specimens—be they rocks or rotifers, one tebrates, Meiobenthos, Molluscs, Arthropods, Worms, specimen or one thousand—are stored in containers, Lower Invertebrates, and Higher Invertebrates. The which are traditionally labelled to show the contents members of the committees are selected on the basis (taxon), date and locality of collection, and other of recognized competence and overall familiarity with pertinent information. Containers, each holding a the community of scientists working in their area. specific kind of specimen, are therefore the items to Each member serves for a period of one to five years, be inventoried, and the container label can serve as

280 ANTARCTIC JOURNAL meter. Also, approximations of data in these fields SMITHSONIAN OCEANOGRAPHIC SORTING CENTER SI-SOS(-180—lsv. 3-67 PORIFERA may be indicated by the abbreviation "ca" for circa. PR GRAM: USARP DATE: 5 JUN 1965 If only the general locality from which the sample COLLECTOR: USC GEAR: 10 1 BLAKE TR - was collected is known, a Marsden Square number VESSEL: ELTANIN LAT.: 55 O1S-55 07S may be entered instead of coordinates. This facili- CRUISE NO.: 18 LONG.:09915W-09919W tates retrieval of records by the locality of sampling. - DEPTH:38434026Th STATION NO.: 1442 Similarly, if the taxon is of such a nature (e.g., phy- ALIQUOT: NO. OF SPEC.: toplankton) that specimen counts are impossible, a REMARKS: volumetric measure in cubic centimeters can be en- tered. In some instances, only the number of lots Figure 2. Example of the SOSC specimen label showing the kinds (containers) may be known, and the entry would of information entered into the computer inventory for the various reflect only this rough measure of quantity. The taxonomic groups found in a sample. entry for aliquot shows that the specimens were re- moved from a completely sorted sample, or prior to subsampling, or else from a subsample—for example, the source document for the information to be cap- a /8 portion of a plankton sample. The inventory also includes with each unit record tured. For the unit record, a label was designed (Fig. 2) the location of the specimens at the time of data that would contain information in a form readily entry, whether they were processed at SOSC, and the understood by persons unfamiliar with the records date that the entry was made. These items are sup- plied by programmed information inserted auto- system, yet one which would satisfy the formating matically during label preparation. The label data requirements of electronic data processing. With the aid of high-speed automatic typewriting systems, are typed and punched linearly on paper tape in the which also can punch and read symbols on paper following order: Taxon, Program, Date, Collector, tape, the labels are prepared and the data captured Gear, Vessel, Latitude (start and finish), Cruise simultaneously on paper tape, later to be edited and Number, Longitude (start and finish), Station Num- transferred automatically to magnetic tape for stor- ber, Depth (minimum and maximum), Aliquot, and age and retrieval. This process eliminates the need Number of Specimens. With the automatic insertion for the more commonly used punched cards, thereby of code symbols to indicate the start and finish of reducing both labor requirements and the probability each field, the computer can distinguish and store of introducing errors by repeated manual transcrip- each type of data in the appropriate data category. Any remarks typed on the labels are excluded from tions of data. the computer entries.3

The Label Entries—Basic Data Programming Catalogs The records system is oriented towards mainte- nance of inventories of taxonomic groups of speci- Anticipating that the data bank on antarctic speci- mens along with the data essential to their scientific mens alone will eventually contain enormous num- value. The types and quantities of data most neces- bers of records, some information on the labels is sary for taxonomic studies were selected with this coded and stored in shorter forms than those appear- fact in mind and with due consideration to the cost ing on the label. Such coding is accomplished auto- of electronic data processing. The label in Fig. 2 matically during computer input. We attempted, illustrates the major data categories—or fields— however, to devise coded forms that were either com- established for the system. monly known abbreviations or easily learned nota- The fields for Program, Collector, Vessel, Cruise tions. There is often little or no difference between Number, and Station Number identify the collection the information on the label and that stored in the and serve as the sequence key for data storage. The data hank. For example, USARP becomes AUSR, fields for Taxon and Number of Specimens show the kind and quantity of organisms, and the remaining fields include the basic data for the sample. It should be noted that the Marsden Square grid system, designed to cover the earths surface and widely used by The fields for Latitude, Longitude, and Depth may other institutions including the National Oceanographic have only one entry, or they may include data for Data Center, provides poor precision in the polar regions. the start and finish positions and minimum and maxi- The most commonly used remarks give the general locality mum depths of the sample. Coordinates may be re- of sampling, primarily of near-shore or intertidal collec- corded to minutes, seconds, and tenths or hundredths tions, for the benefit of specialists who receive the speci- of minutes. Depths are recorded to the nearest whole mens for study.

November–December 1969 281 while USC (University of Southern California) re- cimens prior to their identification to the species level. mains USC. Because of the need for coding, four The second file contains records on specimens which catalogs were developed, each giving the appropriate have been identified to species. Programming for label entries and the corresponding coded forms for the first file was completed in February 1968, and storage and retrieval. The catalogs are for the label the file contained approximately 60,000 records by fields Program, Collector, Vessel, and Gear. Only July 1969. Programming for the second file, the gear, of which few types have commonly accepted Species Inventory, began in September 1968 and the and distinctive abbreviations, are numerically coded. data bank was begun in January 1969. Because the For ease in sorting and scanning, however, the basic information recorded in the two files is basically the type of gear is indicated in the coded form by suf- same, the Species File being distinguished primarily fixes, e.g., P=planktonic, B=benthic, C=cores, etc. by the inclusion of specific identifications, the two Although the scientific names of natural-history files can be readily correlated to retrieve all informa- specimens are often quite lengthy, abbreviations are tion on any recorded parameters of interest. For ex- generally recognized to be undesirable: The similar- ample, all records on a taxonomic group such as ity of some scientific names would lead to ambiguities polychaetes can he extracted readily from both files, or errors if abbreviated. Such names as I)ecapoda, as can all records on organisms found in a given for example, are equally applicable to more than sample or geographic area. one group: without further qualification, Decapoda could refer to some types of crustaceans, molluscs, or algae. The system, therefore, requires the use of com- Processing an Eltanin Plankton Sample plete scientific names with any necessary qualifiers to distinguish each taxon unequivocally. To illustrate how the files are established and main- In the natural sciences, many different hierarchical tained, including updating of records, and how data classifications of plants and animals exist and, by the are retrieved, we will explain the sequence of data dynamic nature of the sciences, others will enter the roccssing for a hypothetical plankton sample taken aboard USNS literature in the future. Meanwhile, records systems Eltanin. For ease of discussion, we require an arbitrary classificatory scheme that will will follow the records on chaetognaths collected in meet present needs. The classification we adopted the sample through to the point of species identifica- was established empirically on the basis of pertinent tion and final deposition in a museum. literature, requests received from specialists for parti- When the hypothetical sample is first taken aboard cular groups of organisms, specimen records from the vessel, pertinent data are recorded by the collec- museums, past SOSC sorting records, and personal tor, usually a member of an institution in charge of communications with specialists. Although the taxo- the plankton sampling program. The sample and nomic catalog as established covers the spectrum of related data are sent to SOSC for sorting and dis- the animal and plant kingdoms, the complexity and tribution of the taxonomic groups to specialists. A degree of subdivision included for any one taxonomic technician at SOSC examines the sample, or a suit- unit varies widely depending on the expected needs able subsample, and classifies the organisms into in that unit. For example, the small phylum Pogono- homogeneous taxonomic groups. Suppose he finds 20 groups including 35 chaetognaths; he then prepares phora is not further subdivided, while all orders and a list of these groups. After determining and format- families of fishes are included under Pisces. The scientific names used, therefore, range from kingdom ing the sample data for computer input, using the programming and taxa catalogs, he requests the to generic levels, while miscellaneous categories to cover all material encountered include such terms as necessary 20 labels from the Records Section. At this point, automatic data processing begins. unknown, unsorted, wood, rocks, and residue. Much of the information needed on each of the The taxa catalog is quite open, i.e., additions can be made easily at any level. Although changes in the 20 labels is repetitive; only the taxon and number of Specimens are variable. Using a CDC-2816 electronic basic structure which would alter the level of exist- system (Fig. 3), the repetitive data (typed once and ing categories impose a more serious problem in that punched simultaneously on a paper tape which is a portion or all of the data bank also would require edited by the operator) are entered automatically on revision, such changes are not impossible, cost being the prime inhibitor to numerous alterations. the labels via a tape reader and the typewriter. The CDC-2816 operations are specially programmed to control the data input procedures including tab set- tings, carriage returns, insertion of special code sym- The Master Inventory Files bols, etc. At the variable fields for taxon and number The system comprises two Master Files. The first of specimens, control is returned to the operator for file, the Inventory of Samples, includes data on spe- manual typing. Each character that is typed auto-

282 ANTARCTIC JOURNAL ple. Similarly, the information entered in the coded fields is now in code form as a means of reducing storage needs and restricting the record to two lines of printout. Also, the location of the specimen is shown as SOSC and date of entry is recorded for in- house bookkeeping. Invoice preparation, too, is part of the data pro- cessing system. Some time after sorting, the chaetog- naths will be sent from SOSC to a scientist at, say, Scripps Institution of Oceanography in California. At the time of shipment, an invoice will be prepared that probably will include the chaetognaths from many samples taken during the cruise. In a manner similar to that for label preparation, a paper tape is generated containing the identifying information (PCVCS and Taxon), the new location, and date of shipment. The data on the invoice tape are sent to the computer and compared with the records in the Master File. When a match on the key fields occurs, Smithsonian Institution Photo the former location of SOSC is changed to Scripps. Eventually, the scientist at Scripps will complete Figure 3. High-speed automatic typewriters used for label prepa- ration simultaneously produce a punched paper tape containing the study on the chaetognaths. Let us assume that data for computer entry. the 35 specimens which left SOSC included three species. Through contact with the specialist, the genus and species names will be obtained. The origi- matically and manually is punched simultaneously nal record in the Sample Inventory will be deleted on a paper tape. Special code characters that are and the new records on three species, including entered on the tape indicate to the computer when counts, if known, will be entered into the Species new records are starting, the beginning and ending Inventory. If the specialist has finished with the spe- of each field, and the end of each record. cimens, representative lots may have been sent to one The computer tape for this sample will be added or more museums for final deposition; if so, the new to those for other samples, and with a header tape locations for each group of specimens will be entered which indicates that the data originated at SOSC, into the file. Indications of type specimens and cata- the resulting reel of tape is sent to the Information log numbers are included. Systems Division of the Smithsonian Institution. Data entry into the Species Inventory is accom- There, the paper tape is machine-read and the data plished with the use of keypunch cards rather than transferred to magnetic tape. During the transfer, punched paper tape. Usually, species listings received the computer (a Honeywell 1250) is programmed to from the specialists or from publications are not edit the various fields of data. The first edit program arranged suitably to take advantage of the benefits checks each field for completeness as indicated by of high-speed automatic entry of data using the the beginning and ending of field and record codes. CDC--2816. That is, the most common types of It also checks the appropriateness of data in some species document will list the sample numbers for fields in terms of the length of the data recorded. each species found in the collection rather than the Records which do not pass this edit are flagged on species found in each sample. Only the latter type the printout sheets which are then returned to SOSC of document could advantageously be processed on for correction. During the second edit, the entries for automatic, tape-punching equipment, as the data for Taxon, Program, Collector, Vessel, and Gear are each sample would be repetitive. checked for conformity with the programming cata- logs, and errors are flagged and printed on a reject list. All "clean" records are then stored in the Master Processing Records on Collections at Other In- Inventory File. The taxa in the hypothetical sample stitutions are now listed alphabetically under the major filing Only a portion of the natural-history collections sequence of Program, Collector, Vessel, Cruise Num- taken from the Antarctic are processed and distri- ber, and Sample Number (PCVCS). Before reach- buted by SOSC. These collections have consisted ing the Master File, during the second edit proced- primarily of planktonic and benthic invertebrates ures, the coordinates of the sample have been used collected during the Eltanin cruises. For other ant- to calculate a Marsden Square number for the sam- arctic collections, we must obtain records from either

November—December 1969 283 the original collector who may be studying the ma- has been programmed to extract only records with terial, or from another curator. Many of these collec- specified parameters. For example, if a specialist re- tions have been inventoried, including pelagic and quests information on the whereabouts of all chaetog- benthic fishes at the University of Southern Califor- naths collected during the months of November- nia; invertebrate specimens, birds, and cryptogams January below 550 m south of latitude 63°S. in the at the U.S. National Museum; part of the entomol- Pacific Ocean, the printout will include only records ogical collections at the Bernice P. Bishop Museum; having these characteristics. bryophytes at the New York Botanical Gardens; cephalopods at the Institute of Marine Science, Uni- versity of Miami; and invertebrates collected by Conclusions Lamont Geological Observatory biologists aboard Vema which are now being sorted and distributed The described data processing system by neces- to specialists from the American Museum of Natural sity and intent will neither fulfill all needs of the History. scientific community concerned with natural-history The processing of records obtained from other in- collections, nor all those of any given scientist. It does stitutions is quite similar to the methods used for have, however, the potential of becoming a valuable specimens which have passed through SOSC; how- scientific tool aside from the master inventory serving ever, actual labels are not required. If the specime1s as an index to antarctic specimens, the latter being have been identified to species, the records will be unique in its scope, versatility, and subject matter. entered in the Inventory of Species file. If identifica- Eventually, the expected data bank and the speed tion has been made only to a higher level, the records with which computers can extract, transpose, and will be entered into the Inventory of Samples file. manipulate data predicts a time when some of the The data in either event are essentially formated, tediousness of traditional natural-history studies can stored, and retrieved from both files in the same man- be left to electronic and mechanical devices. One ap- ner. Retrieval of information on any specimen plication, for example, could be the construction of group, sample number, or collection type can include an atlas showing the horizontal and vertical distribu- collation of all available records whether the or- tions of all antarctic copepod species. With present- ganisms are at SOSC, have been shipped, or have day XY plotters and the growth of the data bank, always been at some other institution. a qualified scientist would have access rapidly to such an atlas, freeing hirn for other pursuits and thus possibly increasing the rate of acquisition and Data Retrieval the depth of knowledge of the worlds biota. Acknowledgements. We wish to acknowledge In designing data retrieval methods, we recognized the contributions of ideas from many of the staff of that many of the demands on the system would re- the Information Systems Division of the Smith- quire printouts or other manipulations of large seg- sonian Institution during discussions and testing of ments of the data bank. To meet such needs, the the records system. The final design and program- Master Files are sequenced on the key of PCVCS as ming of the system was accomplished largely with discussed earlier. Under this sequence, the taxa are the assistance of Mr. Frederick R. Krazinsky, former- listed alphabetically for each sample. Any given ly with the Smithsonian and now serving as a con- record, therefore, can be readily found in a list of sultant to the SOSC Antarctic Records Program. Mr. the records by scanning in the same manner as one Krazinsky also verified the technical accuracy of the would consult a telephone directory. Also, since a system description presented here. Also, we wish question commonly asked is, "Are specimens of taxa to acknowledge the contributions of Mr. Howard X available at SOSC and in what quantity," the data Balduc who, under contract, programmed significant banks can be resequenced in order of locations and portions of the system. Encouragement in the de- taxonomic groups, and the resultant printout again velopment of the system, and requirements from consulted as a directory. The two directories are current biological research were provided by Dr. periodically updated and produced as needed. G. A. Llano, NSF Program Director for Antarctic As the volume of records in the data bank grows, Biology. however, such "directories" will become cumbersome, and specific questions will yield to visual scanning References with greater and greater difficulty. Another concern Llano, G. A. 1965. United States Antarctic Research is the cost of excessive operation of the computer Program in Biology. BioScience, 15(4) : 254-258. while printing the entire data bank when only a Sandved, K. G. 1965. Stateside services for USARP scientists. BioScience, 15(4): 258-261. small portion of the records may be pertinent to the Wallen, I. E. 1963. Oceanography at the Smithsonian question. To alleviate this problem, a query system Institution. Nature, 199(4900) : 1227-1230.

284 ANTARCTIC JOURNAL