M. A. EL BATAWI (World Health Organization, Geneva, Switzerland): The opening presentation made an attempt at defining occupational disease. The World Health Organization has published a book on the interaction between work and health demonstrating the different levels at which work may influence health. For example, the occupational disease of lead poisoning cannot be caused by carbon monoxide; it has to be caused by lead exposure. That is a causeleffect relationship-one cause, one effect. Other such relationships include carbon monoxidecarboxyhemoglobin, asbestos-asbestosis, and silica-silicosis. These can be called the “hundred percent” relationships. Beyond those few relatively straightforward examples, however, we consider many of the diseases that affect the working population and human beings as a continuum: a person has his work; he interacts with so many pollutants in food and air and water; he undergoes stress at work and stress at home; and all these things add up to the well-named diseases of multifactorial causes. Now, I am distinguishing between the terms “occupational” and “work-related,’’ which implies a partial relatedness. Here we do not find causes, we find risk factors. And there is a difference between a cause and a risk factor. For example, if you discuss chronic obstructive pulmonary disease (COPD) or occupational asthma, smoking plays a very important role in causation as does pollution in the work environment and irritants from the atmosphere. Multiple factors play a role in COPD, yet when you diagnose a case of chronic obstructive pulmonary disease in a person exposed to dust, you are faced with the question of compensation. Do you compensate this person and declare the condition an occupational disease? Or is the disease only partly work-related? To what extent is it related to occupation? Epidemiologists are facing a crucial and a very important matter in assessing the role that work may play in the causation of such general diseases as the low back pain syndrome, hypertension, cardiovascular diseases, and stress-related conditions. Unless you consider the person in his or her totality, you will never be able to pin down the causative factors or the risk factors towards which you aim to direct the control measures. IRVING J. SELIKOFF (Mt . Sinai School of Medicine, New York, N . Y . ) : Dr. Fowler, of the 100,000 or so workers now in these high-tech industries, have baseline surveillance mechanisms been set up to see what the problems are going to be? BRUCE A. FOWLER (University of Maryland Medical School, Baltimore, Md. ) : There is a great concern in the semi-conductor industry about these issues. They are moving very rapidly in the direction of setting up good surveillance procedures. In this case, it is the large manufacturing companies that are very concerned about their workers, who represent a highly trained work force, and it is in everyone’s best interest that these workers be maintained in the best of health. However, the greater concern is with the smaller companies, those with fewer than 50 or 100 employees. These workers may not get enough attention and may fall through the cracks of a surveillance system that is being set up mainly by the larger companies. MORTON CORN (The Johns Hopkins University, Baltimore, Md. ) : As you may know, at the Johns Hopkins School of Hygiene and Public Health we have designed and proposed epidemiologic studies for certain of the larger semiconductor companies. I would not want the impression to be left that the exposures in that industry at the leading edge of chip manufacture are of a magnitude that would

HARRY E. MORTON (University of Pennsylvania, Philadelphia, Pa.) : I propose first to direct a few questions to Edward. I t is a great problem to do serologic testing when PPLO cannot be made to grow in liquid media or if they cannot be grown in sufficient quantities to produce an antigen. Can the sensitivity of your growth inhibition test be used for serologic indentification of strains of PPLO that will grow only on a solid medium? Would this be a reliable method, and how would it compare with the complement fixation and hemagglu tinat ion reactions? D. G. FF. EDWARD (The Wellcome Research Laboratories, Beckenham, Kent, England) : I use this test more and more for serologic investigations because it eliminates the difficulty in preparing antigen, although it has the disadvantage of being more expensive with serum. There is much still to be learned about the test, which has been used only to examine a few species of the group. It was used extensively by me for human genital strains from which it is notoriously difficult to obtain good agglutinating suspensions. With the inhibition of growth test more than 90 strains were identified serologically with no technical difficulty. I n the serologic identification of strains the agglutination test has the disadvantage that there are minor differences between strains, so that one speciesspecific antiserum may not agglutinate all strains. For Mycoplasma hominis the inhibition of growth test had the advantage that it was less specific, and one serum inhibited the growth of all strains of the species. Subsequent experience has shown that this may not be true for every species, and there is a suggestion that the inhibition of growth test may sometimes be more specific than the agglutination test. I n preparing antisera in rabbits, growth-inhibiting antibodies do not develop at the same time as do the agglutinating antibodies. Sometimes they appear first; more often the rabbit has to be inoculated repeatedly until they develop. I should be interested to learn whether inhibiting antibodies can be demonstrated in animals recovering from natural infection. I notice that many workers prepare their antigens for inoculating rabbits in media enriched with horse serum. I consider that there is a great danger in doing this, owing to the risk of producing antibodies to horse serum. Before starting this type of work I consulted serologists and they told me that the results would always be open to criticism if horse serum were used. I have, therefore, always enriched my media with rabbit serum. Originally, with strains that grew only poorly in rabbit serum, I spent weeks and months adapting them to grow in rabbit serum media. Eventually I discovered that these strains would grow equally well in a rabbit serum medium as they did in a horse serum medium if cholesterol was added. As it is easy to grow a strain in a medium from which horse serum is eliminated, I strongly advocate that this should be done in order to avoid the criticism that the results may be vitiated by the presence of antibodies to horse serum. In regard to the papers on tissue culture, I can report that I examined ColI mention a general point on the preparation of antisera in rabbits.