ANALYTICAL CHEMISTRY

---

ANALYTICAL CHEMISTRY WALTER J. MURPHY, Editor

what you will. Harry Levin of the Texas Co. pointed out that the petroleum industry has accomplished much through API and ASTh4 to develop sound statistical methods for sampling and that these procedures are readily accessible. Other industries are not so fortunate. Perhaps there is a real need for a rather ambitious program of preparation of bibliographies-a very basic one identifying and describing the literature available that is fundamental t o most, if not all sampling problems, and also a host of more specialized ones dealing with specific fields. Is this a project for the ACS Division of A4nalytical Chemistry possibly with the cooperation of the Division of Chemical Literature? -4 big gap still separates the mathematician or statistician and the analyst, How to bridge it requires thought and then much well-planned action.

most popular session of the recent %day Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy naturally vias the informal roundtable discussion on sampling, with RI. G. Mellon of Purdue serving as moderator. The subject is one of direct concern to every analyst and the packed Pittsburgh Room of the Hotel William Penn attested to the strong desire of the profession to know more about sampling procedures. Sampling is very much like the weather. There is much talk about developing more scientific methods, but comparatively little is being done to achieve such an objective. The approach in the discussion was largely by industry or field. Well-known specialists in ferrous and nonferrous metallurgy and the pharmaceutical and petroleum industries were joined by experts in industrial hygiene and mater, sewage, and sanitation. Each panelist briefly described sampling procedures in his particular field and recounted some of the baffling difficulties plaguing the field. From the more formal presentations of the speakers and the questions and answers, it was obvious that both procedures and unsolved problems vary widely from field to field. There are only relatively few basic concepts in common. These are not too well understood. B. L. Clarke of Merck drew quite a few chuckles from the audience when he somewhat facetiously remarked that the most accurate sample is 100% of the content of the shipment. Sampling as a practical operation constitutes a compromise with the impractical optimum. How far to compromise is the $64 question. T o what degree one compromises involves such considerations as the degree of accuracy that is necessary or desirable and the amount of time that can be justified in each instance. Obviously, sampling a shipment involving hundreds of thousands of dollars requires, from a practical viewpoint, more accurate sampling than one involving a fen- hundred dollars. A product employed in treating disease or a food product must be more carefully sampled than one involving no health hazard. All too frequently, however, sampling is haphazard and inaccurate in analytical work where actual analysis of the product is carried to the third decimal point. Is there any justification for this? The obvious conclusion reached by the audience was neither new nor particularly startling. We need to know more about statistical techniques, statistical design, call it T H E

’

Something Old, Something New EORGE D. BEAL,director of research of Mellon ‘Institute, in his keynote address a t the Pittsburgh Conference, posed a question that tantalizes many analysts. “Khen two scientific disciplines are as closely mingled as are chemistry and physics,” said Dr. Beal, “an ever-present question is \Then to reject the old and accept the new.” Such a question, however, faces all of us in our daily lives, as is indicated by the folloTving quotation from Dr. Beal’s address. “How to rid oneself of the domination of old ideas and methods becomes a problem, especially when those practices and habits hare been impressed by early teaching, and have become precepts because of the faithful service given to them by the older persons whom one has admired and emulated.” We are in complete agreement with Dr. Beal that because of the phenomenal developments in our profession during the past decade or so, the present-day analyst has an especially difficult task of deciding what to retain and what to discard in favor of newer methods and physical techniques usually defined as instrumentation. The decision is not always an easy one t o make, although no real difficulty should be encountered. Dr. Beal described his address opening the conference as the reminiscences of an “elderly man.” His legion of friends and admirers will disagree with the implication. We drank in every word of his talk. We want to share his thought-provoking description of a way of mental approach for men of science toward the fast-changing world of science. The full text will appear in May.

441