. 2004. Leaf development in two-row spring barley under long-day and short-day field conditions. Can. J. Plant Sci. 84: 477-486. To gain an understanding of the adaptation of Australian and Canadian barley (Hordeum vulgare L.) cultivars to the environments of western Canada and Western Australia, phyllochron and number of leaves on the mainstem in eight cultivars of two-row, spring barley were examined when sown at two dates in two locations. The locations were a short-day environment at Northam, Western Australia, Australia in 1997 and a long-day environment at Lacombe, Alberta, Canada in 1998. At each location highly significant relationships between leaf number on the mainstem and thermal time were found (r 2 > 0.94). Using linear estimates, the phyllochron of barley under short days was longer than under long days and was correlated to time to awn emergence. Later sowing shortened phyllochron under short days, but generally not under long days. Error messages from the linear regression analysis suggested that residuals were not random for all cultivars. Bilinear models were fitted to those datasets. Bilineal responses were observed under both short and long days, being independent of cultivar, date of seeding, final leaf number, phenological development pattern and time to awn emergence. The occurrence of a bilinear response was also independent of any ontogenetic events. The change in phyllochron occurred between leaves 4-7 at Northam and between leaves 6-9 at Lacombe. The leaf number at which the phyllochron change occurred was positively related to final leaf number and time to awn emergence. The phyllochron of early forming leaves was positively related to time to awn emergence and shorter than later forming leaves. Leaf emergence patterns in spring barley under both long-day and short-day conditions may therefore be linear or bilinear. ) s'adaptent aux conditions de croissance dans l'ouest du Canada et de l'Australie, les auteurs ont étudié le phyllochrone et le nombre de feuilles sur la tige principale de huit cultivars d'orge de printemps à deux rangs semés à deux dates différentes, à deux endroits. Les emplacements ont été retenus pour leur brève (Northam, Australie occidentale, Australie, 1997) ou longue (Lacombe, Alberta, Canada, 1998) photopériode. Aux deux sites, les chercheurs ont noté une relation très significative entre le nombre de feuilles sur la tige principale et le temps thermique (r 2 > 0,94). Selon les estimations linéaires, le phyllochrone de l'orge est plus long avec une brève qu'avec une longue photopériode, et est corrélé avec le temps écoulé jusqu'à l'apparition des barbes. Un ensemencement plus tardif raccourcit le phyllochrone dans la zone à brève photopéri-ode, mais pas dans celle à photopériode longue, en général. Les messages d'erreur issus de l'analyse par régression linéaire laissent croire que les valeurs résiduelles ne sont pas toujours aléatoires. Les bases de données ont été appliquées à des modèles bi-linéaires puis on a vérifié les réactions avec une brève et une long...