An analysis of slaughter-house returns, covering over 73,000 cattle slaughtered during three weeks in June, 1942, reveals evidence of heterogeneity and interaction in respect of all the five categories involved: time, area, class, origin and degree of infestation with liver fluke. By heterogeneity is here meant that the differences between the various sub-divisions of any one category are too large to be ascribed to chance; by interaction is meant that the relative proportions between the subdivisions under any one category do not remain steady (within the limits of random variation) over the co-ordinate subdivisions of a second category. This makes the data difficult to interpret clearly.