The characteristics of the upper first internode of flowering stems of Eriogonum infla- tur have been regarded as indicators of varietal differences within the species: var. inflatum, with inflations; var. deflatum, without inflations. Inflations also have been claimed to be caused by galling insects. Studies on three populations in central and northwestern Arizona indicate that each individual passes through a stage without inflated stems before producing inflated stems, indicating that these varietal distinctions are invalid. Also, inflations developed without insect attack, and a population in northwestern Arizona has large inflations in the total absence of insect attack. Eriogonum inflatum Torr. et Frem., wild buckwheat or desert trumpet, is a common component of the native flora in many areas of the arid southwestern United States and northern Mexico. Occurring in the Mohave and Sonoran deserts and adjacent uplands, mostley below 1,200 m elevation, the scapose perennial has a basal rosette of leaves and branched flowering stems up to 1 m tall, with striking inflations found in the upper portion of the lowermost internode of many plants, and often in the second and third internodes of larger plants. There is some doubt about the validity of the varieties E. inflatum var. inflatum and deflatum Johnst. (Stokes, 1936; Stone and Mason, 1979), although a recent monograph treats them as biological realities (Reveal, 1976). While the literature generally has regarded the inflated stems as an intrinsic character of the species, Stone and Mason (1979) have argued that these stems are induced by insect attack and should be regarded as insect galls (caused by pyralid moth or mordellid beetle larvae). This study was initiated to attempt to understand the nature of inflations, partly because the arguments made by Stone and Mason (1979) were not com- pletely convincing: (1) In one population they reported that 97% of inflated stems showed insect damage(p. 78), but the question of why 3% should have inflated stems but no insect damage was not addressed. (2) The authors noted that in greenhouse studies neither seedlings nor transpanted stock developed inflations and concluded that in the absence of insects these did not develop (p. 79). The alternatve hypothesis that conditions in the greenhouse were not sufficiently favorable for development of inflated stems was not considered. (3) the authors observed that larger inflations usually contained larvae whereas smaller inflations were damaged but without larvae, concluding that the longer a larva was present the larger the inflation would become (p. 80). However, the alternative hypothesis was not evaluated that larvae select preformed inflations and remain in them according to the amount of food present. (4) The authors also argued that inflations on internodes above those attacked were induced by the larvae in the lower internode (p. 80-81). This would be a most unusual effect of a galling insect without an apparent adaptive value to the insect. (5) Stokes (1936) had suggested that the inflations failed to develop in extremely dry conditions, but this possibility was not examined by Stone and Mason (1979).