The ability of conidia of Leveillula taurica from one host species to infect other hosts belonging to different families has been demonstrated in the Sudan. Several cultivars of Egyptian cotton ( Gossypium barbadense ) and of American cotton ( G. hirsutum ) became infected when artificially inoculated from diseased leaves of Euphorbia heterophylla (an ubiquitous weed in Northern Sudan) or Faba bona . Cross-inoculations in the reverse direction were equally successful. On the other hand, conidia from diseased leaves of Abutilon figarianum failed to cause infection of any other host. Conidial measurements made from different hosts collected at different localities showed that those obtained from A. figarianum were consistently smaller than the rest. It is believed, therefore, that this ‘strain’ must be a different variety or even species of Leveillula . The washing of healthy leaves of E. heterophylla , after inoculation with conidia of L. taurica (at Shambat) had little or no effect if delayed for more than 3 hr. after inoculation. The treated plants, kept at 100% r.h. , showed no reduction (or increase) in percentage infection when compared with treated plants in the ‘free’ atmosphere. Similar results were obtained on young barley leaves inoculated with conidia of Erysiphe graminis at Rothamsted Experimental Station. Since penetration by the infection hypha of both parasites (and of the other species of the Erysiphaceae) occurs normally 24–48 hr. after inoculation, the resistance of the conidia to washing off the leaf surface must be due to ‘physical’ adhesion prior to penetration. The infection and development of the disease under constant saturation being quantitatively similar to that under normal climatic fluctuations tends to refute the opinion that powdery mildew infection and intensity are reduced by high humidities. Despite the varying climatic conditions of the Sudan fresh conidia of L. taurica were able to germinate under a wide range of humidity (0–100%). Germination was best at 85–100% r.h. with long germ tubes and turgid conidia, but was poor at lower humidities (0–30%) with short germ tubes and shrunken conidia. Conidia submerged in water gave poor germination; better germination was obtained from floating spores. The effect of humidity and of water tested on conidia of E. graminis from barley (at Rothamsted), gave similar results except in that they were more or less as sensitive to humidities below 95% r.h. as the conidia of E. cichoracearum, E. umbelliferarum and Sphaerotheca fuliginea tested in the Sudan.