This paper analyzes an aberrant group of echinoids in terms of constructional morphology, i.e., as modification of an established “Bauplan” by a set of new functional and morphogenetic constraints and possibilities. The characteristics of sand dollars (flat test, spine differentiation, branched food grooves, lunules) are related to a particular combination of burrowing and sieve feeding in sandy sediments. It has independently evolved from less specialized Clypeasteroids in at least three lineages (Scutellina, Rotulidae, Arachnoididae), which have solved inherent problems differently (sutural interlocking; growth patterning of food grooves and canal systems; lunule formation; weight belts). These three groups have radiated in different degrees due to their different palegeographic histories.

The kinetic model of taxonomic diversity predicts that the long-term diversification of taxa within any large and essentially closed ecological system should approximate a logistic process controlled by changes in origination and extinction rates with changing numbers of taxa. This model is tested with a new compilation of numbers of metazoan families known from Paleozoic stages (including stage-level subdivisions of the Cambrian). These data indicate the occurrence of two intervals of logistic diversification within the Paleozoic. The first interval, spanning the Vendian and Cambrian, includes an approximately exponential increase in families across the Precambrian-Cambrian Boundary and a “pseudo-equilibrium” through the Middle and Late Cambrian, caused by diversity-dependent decrease in origination rate and increase in extinction rate. The second interval begins with a rapid re-diversification in the Ordovician, which leads to a tripling of familial diversity during a span of 50 Myr; by the end of the Ordovician diversity attains a new dynamic equilibrium that is maintained, except for several extinction events, for nearly 200 Myr until near the end of the Paleozoic. A “two-phase” kinetic model is constructed to describe this heterogeneous pattern of early Phanerozoic diversification. The model adequately describes the “multiple equilibria,” the asymmetrical history of the “Cambrian fauna,” the extremely slow initial diversification of the later “Paleozoic fauna,” and the combined patterns of origination and extinction in both faunas. It is suggested that this entire pattern of diversification reflects the early success of ecologically generalized taxa and their later replacement by more specialized taxa.

The stratophenetic approach to phylogeny reconstruction, developed in numerous recent papers by Gingerich, has its strongest data base in comprehensive collections from early Eocene rocks in the Big Horn Basin of northwestern Wyoming. Other examples have been criticised for lack of well-documented collection data, making suitable testing of stratophenetic interpretations of evolutionary modes virtually impossible.

Study of Hyopsodus from the middle Eocene Bridger Formation of southwestern Wyoming provides new collection data. Adequate stratigraphic documentation is available, so Gingerich's graphical presentation can be essentially duplicated. In contrast to the complex branching pattern Gingerich interpreted from his data, Bridger Formation Hyopsodus data seems to show little size change through approximately one million years. This stasis or equilibrium condition, which is considered rare by Bookstein et al. (1978), is the only well developed pattern in Bridger Hyopsodus.

I present a statistical technique for determining the disarticulation sequence of vertebrate skeletons based on the relative numbers of different intact joints in an assemblage of bones. For remains of modern Topi (Damaliscus korrigum) on the margin of Lake Turkana, northern Kenya, the disarticulation pattern is very consistent. This sequence, on dry land, differs from that reported for bovids that have disarticulated in the presence of water. On land the bones first released as single bones are those moved least easily by currents of moving water. The last released are those moved most easily. I develop a model of random scattering that suggests that the rate of dispersion is great at high concentrations of bones and decreases rapidly as the distance between bones increases. This leads to a condition where scattering effectively stops. The area of more or less stabilised dispersion is dependent only upon the mean distance that each random event moves a bone. Tests show that it is unlikely that articulated units themselves are much involved in scattering, and scattering appears to take place throughout the course of disarticulation.

Owls are important consumers of small vertebrates, and because they regurgitate pellets rich in bone, they may be important potential contributors of the concentrated remains of small vertebrates to the fossil record. Owls of three sizes, the large great horned owl (Bubo virginianus), the medium-sized barn owl (Tyto alba), and the small screech owl (Otus asio), were fed a common diet of mice. The bony contents of the pellets were analyzed to determine the amount of bone loss by digestion, bone completeness, and sites of bone breakage. For all three species, only about half the number of bones ingested were recovered in the pellets. Mandibles and femora were most abundant, and pelves and scapulae were the least abundant. Screech owls broke 80% of the cranial and limb elements, barn owls only 30%. Skulls fared poorly in great horned and screech owl pellets, while barn owls returned 80% of the skulls intact, with only the caudal portion of the cranium damaged; barn owls also returned articulated strings of vertebrae and complete paws. These results provide a baseline for the recognition of owls as agents of accumulation of small bones in the fossil record and suggest that the actions of ancient predators may be revealed by species-specific patterns of bone destruction of an assemblage of fossil prey species.

An understanding of the thermal strategies of extant animals is a prerequisite for any useful discussion on the possible thermal strategies of dinosaurs. Of cardinal importance is the identification of the selection pressures that have resulted in the evolution of high body temperatures. Although different selection pressures have probably been operational in different organisms, enhanced muscle power may have been a prime factor.

Many dinosaurs, including the ornithomimids, carnosaurs and hadrosaurs possess cursorial features suggesting high activity levels, which in turn suggest the possession of high body temperatures. Even if it is denied that high body temperatures are a prerequisite for speed, it cannot be denied that muscular activity liberates large quantities of heat, and most dinosaurs, by virtue of their small area to volume ratio, must have maintained constant, and sometimes high, body temperatures. This thermal strategy, described as inertial homeothermy, does not require a high (mammalian/avian) rate of metabolism, and most dinosaurs were probably not endothermic.

The smallest dinosaurs (e.g. Compsognathus) were probably too small to have been able to maintain constant body temperatures in the absence of external insulation. Perhaps they were feathered, like their close relative Archaeopteryx. Presence of feathers in Archaeopteryx is compelling evidence for endothermy.

The largest dinosaurs, the sauropods, had an exceedingly small area to volume ratio, that probably imposed limitations on their rate of heat dissipation. This in turn may have restricted their rate of muscular activity; there are also sound mechanical reasons why they should have been relatively inactive. It is therefore concluded that sauropods probably maintained relatively low body temperatures.

We present a quantitative method for describing how heterochronic changes in ontogeny relate to phyletic trends. This is a step towards creating a unified view of developmental biology and evolutionary ecology in the study of morphological evolution. Using this representation, we obtain a greatly simplified and logical scheme of classification. We believe that this scheme will be particularly useful in studying the data of paleontology and comparative morphology and in the analysis of processes leading to adaptive radiation. We illustrate this scheme by examples drawn from the literature and our own work.

Regular discontinuities in the distributions of morphological measurements may conceivably reveal tempos in evolution. Recently L⊘vtrup et al. (1974) attributed apparent stepwise interspecific variation in mammalian and avian body size to an evolutionary cause: according to these workers, the pattern may reflect punctuation in phylogeny caused by speciation. The nature and appropriate use of data, the choice of techniques for detecting patterns, and the implications of the results of such tests are discussed here in the light of this Swedish study. A technique, which may be used to verify patterns in rose diagrams or a range of ostensibly rhythmic phenomena, is described and applied to new data. Clustering is found at intervals other than those predicted by L⊘vtrup et al., but there is no consistent trend in the values of these interval lengths among data sets.

Experienced workers in a field are unlikely to change their assessment of basic philosophical issues. Einstein for example never accepted the indeterminacy required by the Heisenberg uncertainty principle, despite the urgings of Max Born and virtually all others. Similarly even earlier, some older workers had refused to accept the young Einstein's theories. Now if issues in physics are not open and shut, they are even less so in paleontology. Possibly it is useful to probe the underlying basic philosophical background to our discipline. The purpose of the present article is to indicate that the inductive, deterministic approaches which have been used almost exclusively in paleontology have stringent limitations. For some types of problems, deductive stochastic processes are more interesting. If we can make the distinction of ‘few’ versus ‘many,’ then it is quite easy to consider deterministic explanations for the individual cases (the movement of individual molecules) but to reserve stochastic explanations for the ensembles of events (the movement of molecules too numerous to count).