In the early years of plant virus research, virus detection and identification were based on symptom development. R. E. F. Matthews points out that the reported host range of a virus can be rather meaningless because in many of the published host range studies only positive results were recorded. The extent of the natural host range of viruses varies greatly. There are viruses whose natural host range covers only a single plant. Host range studies for diagnostic purposes are most useful for those viruses with a relatively narrow host range. There are many uncertainties in using test plants to place an unknown virus in an established virus group. Disease symptoms and host range for natural and experimental hosts can often given an indication of what virus or viruses might be involved in a particular disease. Any factor that reduces or increases the susceptibility of a potential host species may affect the experimentally determined host range.