Coming to the two German reports, it is stated by Weber (1903) that _E. palustre_ contains a specific poison for cattle and other ruminants, but sheep and goats are able, owing to their fine muzzles, to separate it in fodder, and hence suffer less. Horses and pigs, he says, seem to suffer very little. Young animals and stock, from districts where the species does not occur, suffer more than those from places where it occurs—the latter appearing to learn early to avoid it. Lohmann conducted feeding experiments with guinea-pigs with _E. arvense_, _E. palustre_, _E. pratense_, _E. sylvaticum_, _E. maximum_, and _E. heleocharis_ (not British). He also fed _E. arvense_ and _E. palustre_ to horses, cattle, sheep, pigs, and geese in considerable quantities for many days on end; and made experiments with aconitic acid on guinea-pigs and horses. The feeding experiments with guinea-pigs showed that of the species named only _E. palustre_ and to a less extent _E. sylvaticum_ are poisonous plants (to guinea-pigs). With the large domestic animals the experiments showed _E. arvense_ to be a harmless plant, and _E. palustre_ to be really injurious to cattle but avoided by other stock. Lohmann considers that the many statements in the literature agree in part with this result, and that the divergent observations may be traced to various causes, among which perhaps an abnormal chemical composition of the weed fed plays a principal part. In this connection, however, the American results must be carefully borne in mind, and _E. arvense_ must not too hastily be regarded as blameless.