sorting all cards of the file arranged according to chemicals is too great to be practical (again, speaking 

 with respect to machines with the speed of those used by the CBCC). Another factor related to this can 

 not be ignored, the vulnerability of the cards to a certain amount of damage with handling and the passage 

 of time. Since their practical "life" is of limited duration even with minimal use, the frequent passage 

 of all the cards of a single file through machines would inevitably demand frequent card replacement 

 due to damage. 



The answer to this problem was to reduce to a practical minimum the number of cards to be 

 machine handled in any given card sort. To do this, the CBCC established eight card files, supple- 

 mentary to the principal file which was arranged according to the Serial Numbers of chemicals tested. 

 For example, for the "Taxonomy Punched Card File", a duplicate was made of every card in the major 

 file (i. e. , the "Serial File") and the cards were arranged according to the symbols coded in Field E 

 (organism, tumor, or pathology tested); in six other files, the cards were arranged according to the 

 coding in Field T-2 (the biological state, quality, or process affected), in Field T-3 (the category of 

 effect of the test compound reflecting practical use), in Field H (the organ affected), in Field I (the 

 tissue affected), in Field J (the host of the test organism), and in Field D (a chemical synergized, 

 antagonized, or used as a standard for comparison in evaluating the effect). More recently, an 

 additional (eighth) supplementary file was established which resembles the Taxonomy Punched Card 

 File just described in that the cards are arranged according to symbols coded in Field E; it differs in 

 that its arrangement is not on the basis of the finest classification unit (i. e. , species of test organ- 

 isms), but on the basis of major classification units (e. g. , bacteria, dicotyledonous plants, arthropods, 

 etc. ); under each of these major taxonomy classification units , all the cards are arranged by Chemical 

 Serial Number, whereas in the principal Taxonomy Punched Card File, the cards are arranged by 

 Chemical Serial Number under each species . 



Thus, in all, there are nine separate Biology Punched Card Files. The nine files are not all of 

 equal size, because certain of the information categories are not represented on every card and, if a 

 card is not punched in a field for which a special punched card file was established, that card is not 

 included in that file. The Serial File and the two files arranged by entries in Field E and Field T-2 are 

 all of equal size, each containing a card for every code line. On the other hand, the files arranged 

 according to tissues and secondary compounds are small enough to be accommodated by a few cabinet 

 drawers. 



It should be noted that, for most of the fields (i. e. , indexes), such as Field A (state of the 

 chemical), Field G (experimental state of the organism), Fields M and N (dose size), and Field S-3 

 (path of administration of the chemical), no special punched card files were established. This is 

 because of the improbability that the entire Serial File would ever need be sorted for that information. 

 In other words, separate punched card files were maintained only for those categories of information 

 most frequently sought and for which a machine sort of the entire punched card file arranged by 

 Chemical Serial Number would be necessary but impractical. For example, if there were wanted all 

 information about tests made on a specified organism or group of organisms, or on a specified organ, 

 or for a specified response, the file search would begin with punched cards bearing the identifying 

 symbol for the organism, organ, or response in question; the cards would be obtained by manual 

 selection from the punched card file arranged according to test organisms, from the file arranged 

 according to the organs responding, or from the file arranged according to specific responses. 



The type of question asked of the CBCC files was seldom so simple as the examples given 

 above, but was most frequently qualified by stipulations of a second or more information categories. 

 For example, a request might be for information about tests made for a specified response on a speci- 

 fied organism; this might be further qualified by stipulating interest in only positive data and in chem- 

 icals effective at doses lower than a specified maximum . No special files of IBM punched cards are 

 maintained by the CBCC with cards arranged according to evaluation of effectiveness (negative or pos- 

 itive or the degree of positive activity) and dose size. These are not categories of information for 

 which a primary search through the entire card file would ever be likely. Therefore, the first step 

 would not be to machine sort the entire punched card Serial File, but to go to the appropriate supple- 

 mentary punched card file, in this case either the file arranged according to test organisms or the file 

 arranged according to responses, or both. At that point, depending on the number of cards which had 

 been found for the specified organism or for the specified response, the decision would be made as to 

 whether the machines could be used to advantage in sorting for the cards meeting the other specifications 

 (data which is positive and at doses lower than a given maximum). It is possible that the cards result- 

 ing from the primary step may be so few that, for the secondary sort, reading and sorting the cards 

 manually would take less time than it would take to carry the cards to the machines, make out specifi- 

 cations for machine operators, etc. 



209 



