J D Myers 1974 Journal The Tower (16) Estancia M[illegible], 35 Kms S of Juanelo by road, Route de Buenos Aires Argentina 21 October (cont'd) up in the 1530 sample. The pattern of usage changed to a much less aggressive one during the afternoon. In some cases I continued to note occasional instances of aggression. However big and large the buffalo on the grid in the afternoon were foraging in directional flocks. # samples with graze # birds present 25- 20- 15- 10- 5- # birds on 17b in a given sample • bufflo • goldens During both morning and afternoon survey, Pluvialis were very territorial. For all samples taken during the day on 17b I have computed a measure of the dispersion of the birds in terms of their degree of clumping to random dispersion. The ratio of λ/μ from Cloyd's article on mean crowding, λ being the mean number of individuals in a unit per individual and μ being the mean/unit area. A value of 1 indicates random dispersion. Values exceeding one indicate a clumped or patchy dispersion and <1 indicate an even dispersion. The following table presents mean values computed for bufflo and goldens during the morning and afternoon. morning 11 samples afternoon 12 samples bufflo 1.81 (0.76 - 5.5) 2.79 (1.1 - 6.67) golden 0.48 (0 - 1.0) [illegible] (0.83 (0 - 1.78)) Note that the samples for which the means were computed cannot be regarded as independent. In a graph of λ/μ = f(twin of sample), one feature stands out. (The graph is not shown here, see grid data book). Bufflo rhythmically move in and out of a clumped dispersion. The period appears to vary, but, eyeballing the graph, there appear to be definite oscillations - alternations between periods when the birds were largely clumped (peaks on the λ/μ graph) and when they were largely dispersed. In the afternoon, bufflo tended to be more clumped than they were in the morning. Recall that observations in qualitative terms noted directional flocks in the afternoon but more