develops gradually—In whose employment was the disease contracted? Clearly in such a scheme of insurance against both accident and industrial disease only specific industrial diseases would be included, i.e. diseases in which the connection with the industry can be clearly established as due to causes inherent in the industry, and traceable to definite materials used. Such diseases as tuberculosis and the effects of dust inhalation (bronchitis, &c.), which as industrial diseases occur only too often, cannot be called specific, because they arise outside the industry and make decision impossible as to whether or not in a particular case the disease owed its origin to the occupation. In order to determine what should be regarded as specific industrial poisons it was deemed necessary to draw up a schedule. For one such list Sommerfeld (in collaboration with Oliver and Putzeys) is responsible, Carozzi of Milan for a second, and Fischer[F] for a third, published in 1910. Those by Sommerfeld and Fischer are constructed in similar fashion—enumeration of (1) the poisonous substance, (2) the industries in which it is made or used, (3) the channel of absorption, and (4) the symptoms produced. Sommerfeld enumerates the poisons in alphabetical order, noting against each the requisite preventive measures, while Fischer adopts a chemical classification, confining himself to general introductory remarks as to prevention. Sommerfeld proposes to limit notification to poisoning sharply defined as to the symptoms set up, such as lead, phosphorus, mercury, arsenic, chromium, carbonic oxide, aniline, benzene, nitrobenzene, carbon bisulphide, and nitrous fumes. This simplifies the obligation to notify, but does not dissipate the fears expressed above as to the difficulty, because in the present development of the chemical industries new substances involving new danger to the persons handling them are constantly being discovered, and thus there can be no finality as to which industrial poisonings should entitle to compensation. And if recourse were had from time to time to additions of new substances to the schedule, reliance would have to be placed on experience with regard to each substance added, and thus the actual individual who had suffered would not benefit. Fischer, indeed, acknowledges that any schedule must be incomplete, and emphasises the fact that continual additions would be necessary; otherwise it would be better to refrain altogether from publication of a list. Such lists may be valuable guides, but no sure foundation for insurance legislation. The only possible way to do this is to give as far as possible a correct definition of the industrial diseases entitling to compensation and, in isolated cases, to leave the decision to the expert opinion of competent judges.