Morphological variation between populations of Zonorchis petiolatum (Railliet, 1900) from seven different host species and two different localities is shown to be considerable in every character examined. It is considered, however, that the populations examined were drawn from the same distribution, and that the morphological evidence was insufficient to justify separation into different species. Statistically it is shown that there are significant differences in the egg size of trematodes from hosts in different families. If, when sufficient material is available to carry out further work comparing other morphological characters, similar differences are found, the question of speciation must be reexamined. Because of the large percentage nonoverlap between the populations when compared by a coefficient of difference, it is suggested that the standard of species determination, as far as helminthological work is concerned, may lie near Amadon's (1949) concept of subspecific distinction. This, expressed in terms of coefficient of difference, would indicate specific difference if the CD value was 1.96 or above. If large populations are being compared a higher CD value (2.18) should be used. Increasing attention has been paid recently to the application of a more modem species concept to parasitological systematics. Sonnebom (1957) pointed out that the present biological species concept, as put forward by Huxley, Mayr, etc., was based solely on cross-breeding plants and animals, which do not represent the majority of all organisms. The present use of the term "species" in the taxonomic and evolutionary sense is confusing. Crites (1962), reviewing the future of morphology as a basis for the identification and classification of parasites, has drawn attention to the urgent need for detailed work on speciation. This is required so that some concept can be formulated to explain evolution of groups of parasitic animals that may reproduce asexually at some stage of their life history, and be self-fertilizing and/or sexually reproductive at another. Possibly an acceptable modem species definition for parasites could then be determined. Speciation work involves population analyses and a study of quantitative characters. Taxonomists working at this level will find statistical Received for publication 25 February 1963. methods indispensable, because the principal object is an accurate quantitative description of a population to characterize it in such a way as to facilitate comparisons. As population geneticists in other groups have shown conclusively that, in sexually reproducing animals, no two natural populations are ever exactly alike, it is reasonable to assume that this will be demonstrated also in the Platyhelminthes. It follows therefore, that a mere statistical difference between populations of a species is of no particular interest. Of prime importance, however, is the quantity of this difference and, if the differences are large enough to justify classification of the populations into different taxonomic categories. Mettrick and Weir (1960) discussed speciation within the genus Inermicapsifer Janicki, 1910 [Cestoda: Davaineidae] and have recently (Mettrick and Weir, 1963) pointed out the implications, within the family Davaineidae, of this variation. Mettrick (1963) revised the genus Riberoia Travassos, 1939 [Trematoda: Cathaemasiidae] after examining the variation between various population groups in this genus. Speciation work on free-moving, free-living populations