A Study of Chemical Deliveries as a Function of Cigarette Butt Length

Abstract It should be emphasized that the major conclusion of this paper is not that any particular parameter is linear on any particular scale, but that there are highly significant differences in slope, intercept and perhaps even linearity between brand groups and even brands within groups. In fact, although several brands do seem to be adequately linear on either arithmetic or logarithmic scales they very clearly do behave differently from each other. There is also quite strong evidence that condensate and nicotine are not linear on the same scale, with condensate being approximately linear on the logarithmic scale and nicotine approximately linear on the arithmetic scale. On the basis of the above it is quite clear that the ranking of brands on the basis of measured value of any parameter will depend strongly on the butt length at which the parameter was measured. It is therefore apparent that international comparison of deliveries requires redetermination of deliveries at the new butt length standard rather than mathematical manipulation using a single functional relationship. The form of a suitable functional relationship for HCN cannot be ascertained from the present study and will require additional experimentation. The statistical design and analysis proved worthwhile in model discrimination for the other three parameters.