Measurements of bodily parts were taken throughout life from captive male and female eastern grey kangaroos of known birth date. For each body measurement a growth curve was fitted, and confidence intervals calculated for determining the age of any new animal. A change of phase in the growth curve was apparent at the time when the young vacate the pouch. A join point in the curve was empirically estimated to occur at approximately 310 days, and the growth curve was represented by a 4-parameter non-linear model consisting of two hyperbolas constrained to pass through this common point. Tables presented permit determination of age from body measurements with corresponding confidence intervals for both males and females at 30-day intervals in the first year and less frequent intervals for the age of 15-36 months. Head and leg lengths proved to be the most reliable criteria for age determination; arm and foot length were reasonably accurate; ear length, tail length and weight were unreliable criteria.

Seven body measurements were taken at regular intervals throughout life from both male and female western grey kangaroos of known birth date. For each sex of three subspecies and for each body measurement a growth curve was fitted, and confidence intervals calculated for determining the age of new animals. As with eastern grey kangaroos, a phase change in the growth curve was apparent at the time when the young vacate the pouch. Join points in the curve for each subspecies of western grey kangaroos were estimated empirically, as three-quarters through the interval between first emergence and final vacation of the pouch by a young animal. The growth curves were represented by a four-parameter non-linear model consisting of two hyperbolas constrained to pass through the common point. Tables presented contrast the ages at which percentile growth values are attained for each subspecies, and also provide examples of the determination of age from body measurements for both males and females, at monthly intervals during their first year. As found for the eastern grey kangaroo, head length proved to be the most reliable criterion, and all measurements of animals older than 2 years were unreliable for age determination.

Apparently once widespread throughout dense thickets in south-western Australia, the tammar is now much restricted in its distribution. On mainland Australia, isolated populations still persist in Western Australia, but in South Australia, where there is little remaining evidence to confirm that it extended beyond Eyre Peninsula, the wallaby is probably close to extinction. All originally recorded populations on five islands in Western Australia remain, but in South Australia all natural island populations, other than those on Kangaroo I., appear to be extinct. Morphometric analyses of crania representative of most known populations provide a means of assessing their relationships. Canonical variate analysis, the derivation of Mahalanobis distances and subsequent calculation of minimum spanning trees supported the existence of affinities within three major regional groups-a group predominantly from Western Australia, a group from Kangaroo and Greenly Is, South Australia, and a group from New Zealand-all apparently related via a population from Eyre Peninsula, presumably representative of a former widespread mainland population. By cranial criteria, feral tammars established in New Zealand are South Australian in origin although probably not introduced from Kangaroo I.

Pryor's rule that mixed stands of eucalypt forest consist of species from different subgeneric groups was tested statistically using data from a vegetation survey of part of the South Coast of New South Wales. The plot data were stratified by environmental regions, and expressed in terms of the subgeneric combinations of the two most abundant tree species. The categories recognized were the eucalypt subgenera Monocalyptus, Symphyomyrtus and Corymbia, plus Angophora and others. The results suggest that: (a) subgenera are characteristic of certain environmental regions; (b) combinations of subgenera are not random; (c) a modification of Pryor's rule is applicable to three of the four regions studied; and (d) in addition, certain combinations of subgenera occur more frequently than expected by chance, e.g. Monocalyptus occurs as the most abundant species, with Symphomyrtus as subordinate, more frequently than the reverse situation. The results accord with recent reviews of eucalypt forest ecology but there are many plots with a composition of three species from the same subgenus. Biological explanations for Pryor's rule must also take account of these exceptions and the tendency for Symphyomyrtus species to be subdominant to Monocalyptus in the coastal region.

Seven body measurements were taken at regular intervals throughout life from both male and female eastern and western grey kangaroos. Evaluation of the reliability of criteria for determination of age and some aspects of the growth models for the two species were presented in earlier papers in this series. In this paper the common patterns and relationships between species in the growth characteristics of their body parameters are described and analysed. Comparison is made between species and sexes of rates of growth and size attained both within the pouch and following vacation of the pouch. Head, arm, leg and foot length were important discriminators, particularly when contrasted in various ways to summarize different body proportions. The insular form M.f. fuliginosus readily separated from the mainland forms, and M.f. ocydromus showed some differences which were related to its longer pouch life. Hybrid animals showed growth patterns intermediate to those of their parents. Sexual dimorphism in patterns ofgrowth was not detected during pouch life but was exhibited by all species after the young vacated the pouch and grew towards their full adult size.