Biology IBM Punched Card Files : 



The coded information of the Biology Code Sheets and Chemistry Code Sheets was placed on 

 IBM punched cards and filed in standard IBM card cabinets. The coded and punched information on 

 these cards may be regarded in two ways. 



First, the entries on the card serve to index the Biology Code Sheet File which contains the 

 information about chemical-biological tests in detail as a written abstract. The advantages of this 

 index (more correctly, a complex of indexes) have already been discussed in the Introduction and will 

 be continued here and in a subsequent part of the Appendix. 



Secondarily, the CBCC considered the information of the Biology Code Sheet File to be coded 

 to such detail that, for some purposes, the information of the Biology Punched Card Files might be 

 regarded reasonably as an adequate substitute for the Code Sheet File. At the same time, it was recog- 

 nized that the coded information on the punched cards was not adequate for all purposes. Details 

 unique to special tests are often not coded for lack of a coding provision and certain information is 

 abbreviated to code only by expressing it as an approximation; for example, the coded dosage would 

 be known to lie somewhere within a coded range. Furthermore, since the CBCC IBM punched cards 

 were actually designed having in mind only their use at the Center in conjunction with the Code Sheets, 

 no provision was made for including the source of information on the cards. Thus, for final interpre- 

 tation, it is often necessary to go to the Biology Code Sheet to which reference is made by the 

 punched card. For reference to the original information source, it is always necessary to go to the 

 Code Sheet. This suggests that a desirable eventual refinement would be the design of some means 

 whereby the punched card would include the information source and even a reproduction of the related 

 Code Sheet Line which the punched card represents. Nevertheless, the IBM Punched Card Files can 

 be regarded, for some purposes, as bearing enough information in code to be used independently of 

 the Code Sheets as a medium for transfer of information. 



Because the Punched Card Files and their use represents such an important aspect of the CBCC, 

 it is appropriate to dwell on their arrangement and the reasons behind that arrangement, in an effort to 

 make clear their significance and to give a general idea of the way they were used. However, it is 

 not possible here to outline all the patterns for use of the punched card files in searching for specified 

 information. The pattern varies according to the information requested and, except for certain more 

 simple requests, some "programming" for sorting of punched cards is necessary. 



Reference has already been made to the Introduction's description of the punching of coded 

 information on cards in terms of indexing the Code Sheet Files in which the information is deposited 

 as a written abstract. The information in the Code Sheet File can be considered as being indexed, 

 not merely by the order by which the punched cards are filed (according to chemicals tested, e. g. ), 

 but according to every category of information coded and punched on the cards (i. e. , according to the 

 chemical tested, organism, anatomy, host, response, etc. ). Furthermore, it has been pointed out in 

 the Introduction that an advantage to using machines and punched cards is that it permits having all 

 these indexes in a single file rather than having a duplicate set of cards for every index, each arranged 

 according to that index classification. For example, even though the cards were arranged according 

 to the chemical tested, if information were wanted about a given organism or group of organisms (e. g. , 

 information on MoUusca), the speed of machines makes possible a rapid examination of the entire file 

 and sorting out those cards punched to indicate MoUusca; the results of this machine sort is identical 

 to the result that would be achieved by having a separate file of punched cards arranged according to 

 code symbols for organisms, from which all the cards referring to MoUusca could be taken manually. 



These basic observations would lead to the assumption that the CBCC needed but a single file 

 of IBM punched cards to index its information collection. Since the Center actually established 

 several files of its IBM punched cards, differing only by the way they were arranged (according to 

 organism, to response, etc. ), the above observations need modifying to explain, first, why a single 

 file was inadequate for the CBCC and, secondly, that the advantages described above apply and are 

 utilized even though the several card files are needed. 



For these explanations, two major factors should be considered, (1) the speed of the machines 

 used and (2) the size of the information collection and the corresponding size of the file of IBM punched 

 cards . Machine sorting for cards on which is punched a specified symbol, from a group of one hundred, 

 or even five hundred, cards is one thing, but sorting by machine from a file of many thousands of cards is 

 quite another. It is not a matter of the machines used by the CBCC not being able to accomplish this, 

 but a matter of speed ; compared to the speed by which cards for MoUusca, for example, can be taken 

 manually from a file in which the cards are arranged according to organisms, the time for machine 



- 208 - 



