appropriate, then, to consider the thoroughness of published abstracts with regard to the chemical- 

 biological data with which the Center was concerned. 



However high the quality of the published abstract, it yet remains by definition a condensation 

 of the original data, frequently omitting most experimental details; for practical reasons, a published 

 abstract is restrained to the briefest summary of test data. It can not be expected to describe all the 

 information comparable to that coded by the CBCC, partly because of the limitations in publication 

 space and partly because abstracters do not prepare abstracts according to fixed rules comparable in 

 detail to those for CBCC abstracting. There is often no real assurance that an abstract identifies all 

 the organisms used or all the chemicals tested, or all the responses demonstrated in a given article, 

 not to speak of doses, paths of administration, evaluation details, etc. 



It would be expected that an agency whose efforts were concentrated on a given type of infor- 

 mation would collect that information more assiduously and in greater detail than would an agency 

 whose literature coverage is either far greater in scope or of a different specialization. Thus, in the 

 case of the CBCC, having established exactly the details wanted for its purposes, chemical-biological 

 test data was abstracted to include meticulously all organisms, all compounds, and all responses from 

 any given chemical-biological information source (to restrict the observation for the moment to those 

 three more commonly published subject indexes). This factor of thoroughness is not implied here to 

 be an advantage offered by the CBCC system or procedure precisely, so much as an advantage resulting 

 from the CBCC's specialization . 



Sometimes observations of biological responses are reported incidentally in an article. In 

 other words, there occasionally occur, and are mentioned in passing, responses not further discussed 

 or explored by the author nor are they included in the article's summary. Whatever the reasons may be, 

 the fact remains that it is not unusual for such "incidental" observations to be made obscurely in liter- 

 ature articles. Many examples will come to mind of "discoveries" made, such as the inhibitory action 

 of penicillin or the herbicidal action of chlordane, effects which had been observed and more than once 

 noted incidentally in the literature- -to be overlooked until favorable circumstances brought about the 

 recognition of the practical significance of the observations. It is not to be expected that published 

 abstracts and indexes would unfailingly include all such incidental observations from each article, 

 nor could the Center claim infallibility in recording all incidental effects in its files. The abstracting 

 policy of the CBCC, however, intended that chemical-biological information from the literature (as 

 well as from other sources) should be thoroughly examined and all effects, including those defined as 

 "incidental" should be included, as the previous paragraph has implied. The fact that the Center 

 frequently omitted coding specifically the many subjective effects described as common "side effects" 

 caused by therapeutic administration of test compounds (headache, nausea, dizziness, etc. ) should 

 not be interpreted to mean that there are ever omitted observations made in the literature representing 

 specific biological responses to test compounds, however incidental they may be to the original intent 

 of the tests. 



The comparative observations of the preceding paragraphs are based essentially on the three 

 criteria by which chemical-biological information is frequently indexed and to some degree cross- 

 indexed by subject in published indexes to literature. 



There is a practical limit to cross-indexing in a published index as well as a limit to cross - 

 indexing in index card files. Indeed, there is a practical limit to all cross-indexing, including that 

 done by a center such as the CBCC which concentrates on a given type of Information. The practical 

 extent of the cross -indexing depends on many factors and is not identical for all purposes, all systems, 

 or all types of data. 



In the case of the CBCC, it has already been explained that every coding field of the Biology 

 Code represents essentially a distinct indexing criterion. Beyond indexing according to the chemical 

 affecting the biological system, the biological system affected, and the response, all test information 

 is indexed according to any hosts (of non-infectious pathology, parasitic organisms, or tumors) involved, 

 any organs, any tissues, the quantities of test compounds needed for given effects, any compounds 

 whose biological effects test compounds influence, developmental stages and the sexes of organisms 

 affected, routes of administration, and so on. 



Experience with published indexes provides the most convincing evidence of the time necessary 

 to search through volume after volume of the published abstracts to literature or through volumes of 

 the literature itself (to which the published indexes have given reference) to find specific chemical- 

 biological information. When cumulative indexes are not prepared, the task includes searching through 

 consecutive volumes of the index. Compared to the use of published indexes to find information about 



215 



