Phylogeny is a succession of ontogenies, and the two have been compared by considering them in terms of rates of size change. In development, the larger the organism, the slower its rate of size increase. In evolution, the rates of size change can be put into three distinct categories: fast, medium, and slow. The fast changes occur over short periods of time (1–10 thousand years) and are as likely to show size decrease as increase. The medium changes occur over longer time spans (5–20 million years) and are predominantly or entirely instances of size increase. The slow changes occur over the entire span of organic evolution and represent the maximum size attained in various phyla, which again show an over-all increase.

For ontogeny, a decrease in rate of size change is correlated with an increase in complexity, an increase in the number of gene actions. For evolution, it is correlated with an increase in the number of genetic changes required of the genome by natural selection in fluctuating environments.

Many peculiarities of growth and development in colonial metazoans are related to the fact that a true colony is a clone, composed of genetically identical individuals reproduced asexually from a single founding individual. In fossil colonies the close relationship of individuals may be difficult to prove, but it must be assumed for analysis to be meaningful.

Intercolony variation is due to the interaction of genetic and environmental controls as in solitary organisms. Intracolony variation may be very limited because of the lack of genetic variation, but microenvironmental differences (situation within colony) may cause enough individual variation to mask the restrictive effect of genetic similarity.

Study of the number of major septa in some Devonian solitary and colonial rugose corals shows significantly less variation within colonies than within “populations” or species of colonial or solitary forms. Greater variation within some colonies may result from either genetic or mechanical accidents. Septal number and diameter are closely correlated in populations and species, but analyses of individuals within colonies frequently show weak or no correlation. Available data suggest: 1) that septal-number variation may be genetically limited in rugose corals and 2) that the general assumption that number of septa is a function of diameter may be in error.

Paleontological studies of organisms with accretionary skeletons should take full account of the implications of accretionary growth. These implications are discussed in this paper with examples from the Brachiopoda.

Descriptions of morphology should recognize the ‘dynamic’ character of preserved structures as embodying a sequence of previous growth stages; ‘static’ descriptions can result in misleading morphological characters and spurious taxonomic distinctions. Analysis of ontogeny in dynamic terms, and particularly the topological comparison of structures of varied form, can help distinguish true from false homologies, and hence evaluate the probability of inferred phyletic pathways. Such analyses can also be used to infer the morphogenetic ‘rules’ that governed the development of particular structures and, hence, to locate major phyletic innovations and discontinuities in structure.

Four distinct classes of functional interpretation are applicable to accretionary structures. Functional analysis of structures developing through ontogeny can help to discriminate between analogical and homological resemblances and to identify functional changes during ontogeny. Likewise, functional interpretation of postulated transformations of structure can be used to evaluate the phylogenies suggested; interpretation of structures that are functionally dependent on absolute size can also help to identify major adaptive innovations and discontinuities.

Deutler's classic work (1926) provides the basis for an analysis of plate growth in echinoids. Deutler showed that the concentric growth rings found within each plate may be used to reconstruct the ontogeny of the entire plate pattern.

New observations of growth-line configurations, particularly in Strongylocentrotus pallidus (Sars), have been used to derive a general model for echinoid growth. The plate mosaic is assumed to be the result of close packing of the growing plates. The shape of a given plate thus depends on its size and position relative to surrounding plates.

The rate of meridional growth of a plate can be shown to change regularly with increasing distance from the echinoid's apical system. Migration of plates away from the apical system thus causes change in the rate of plate growth. The rate of plate migration is strongly influenced, in turn, by the rate of supply of new plates at the apical system.

A mathematical model has been developed from these considerations which, when treated by digital computer (with x-y plotter output), produces ideal plate patterns (including growth rings). The constants in the model may be varied to produce a broad spectrum of echinoid plate patterns.

Tidal cycles are reflected in daily growth-increment sequences in shells of many Recent and fossil mollusks. Living specimens of the bivalve Mercenaria mercenaria were notched at the growing edge of the shell and planted intertidally in Barnstable Harbor, Massachusetts. Shells from two lots, killed at intervals of 368 and 723 days after planting, show the same number of small growth increments as there were days from notching to killing. Superimposed on daily growth record are effects of winter (thin daily increments) and tides (14-day cycles of thick and thin daily increments). Comparison of Barnstable tide record with the first year's growth shows that, for each 14-day cycle, thin daily increments form during neap tides and thicker daily increments form during spring tides. Although tidal patterns are present in subtidal Mercenaria shells, they are rarely as pronounced as in intertidal ones. Spawning patterns differ from winter patterns; they are expressed in the shell by an interruption of regular deposition followed by a series of thin daily increments. Continuous sequences of bidaily patterns, one thick daily increment followed by a relatively thin one, are common in M. mercenaria.

The clearest 14-day cycles of deposition were seen in shells of the bivalve Tridacna squamosa. Each daily neap-tide increment is a simple layer consisting of a dark and light zone. Each daily spring-tide increment is a complex layer consisting of two light-dark alternations separated by a depositional break that is more pronounced than the breaks delimiting daily intervals. Preliminary results of growth-increment counts in fossils show a generally decreasing trend of the mean values of days per lunar month toward the Recent. The Pennsylvanian value is 30.07 ± 0.08, a figure that is in general agreement with those of Scrutton (1964), who counted 30.59 days per month on Devonian corals, and Barker (1966), who reported more than 30 days per month in Pennsylvanian bivalves.

Significant allometry occurs during the ontogeny of every variable in Poecilozonites bermudensis, a Pleistocene-Recent land snail from Bermuda. The adaptive significance of shell allometry may lie in the necessity for preserving a high value of the foot surface/body volume ratio. In the absence of allometry, this ratio must decline as size increases. Three strategies could be used to keep this ratio sufficiently high: positive allometry of foot growth, structural strengthening of the foot, and the development of a foot initially large enough to withstand decline in the ratio during growth. As indicated by ontogenetic changes in apertural shape, a small amount of positive allometry occurs during ontogeny of the foot in P. bermudensis. This is not sufficient to prevent decline of the foot surface/body weight ratio, and I conclude that the strategy of possessing an initially large foot is also used. A simple model of doming, reflecting this latter strategy, is constructed (doming is a major allometric feature of P. bermudensis). In this model, the foot volume/body volume ratio is constant throughout ontogeny in each of two shells, but this value is higher in the more strongly domed shell.

Knowledge of ontogenetic allometry is a prerequisite for understanding the phylogeny of P. bermudensis, for paedomorphosis has been the primary evolutionary event in this taxon. Paedomorphic samples are scaled-up replicas of juvenile shells of the central stock, P. bermudensis zonatus. The degree of ontogenetic retardation in development is the same for all variables (color, thickness, and external shape). Paedomorphosis has occurred several times during the Pleistocene, providing an example of iterative evolution at the infraspecific level. Four paedomorphic taxa are known: P. b. fasolti, P. b. siegmundi, P. b. sieglindae, all new; and P. b. bermudensis (Pfeiffer). They have the geographic distribution (small, peripheral isolates) expected of diverging populations and seem to be genetically distinct entities, not mere phenotypic variants. The most paedomorphic subspecies originated in red soils; paedomorphs did not evolve in times of carbonate-dune deposition. The thin shells of paedomorphs might have been adaptive in the low-calcium environment of red soils. The general significance of iterative evolution at the infraspecific level does not provide an adequate model for corresponding events at higher levels.

The calyx of a Paleozoic camerate crinoid is composed of a mosaic of large, polygonal plates. Ontogenetic size increase results from accretionary additions of calcite to thecal plates. When new calcite is added to the lateral edge of a plate, a new layer is also added over the external surface, obscuring the earlier growth stages of the plate. However, ontogenetic development can be studied by measurement of a growth series. This is illustrated by regression analysis of growth series of two Silurian species of the genus Eucalyptocrinites: E. crassus (Hall) and E. tuberculatus (Miller & Dyer). Growth of the principal plates is isometric; most growth parameters have a very similar mode of development in both species; only a few parameters show any specific difference. A table of correlation coefficients for 51 growth parameters demonstrates that the entire development of the calyx was highly coordinated throughout life, with coefficients almost always equal to or larger than 0.90. Integrated expansion of the cup plates provided a larger cavity for the viscera. The stem increased in diameter to support the increasingly larger theca, as did the diameter of the root system. Growth of the vaulted plates in the upper part of the theca provided a protective recess for the arms, which probably contributed to the evolutionary success of this animal on a widespread basis during the Silurian and Devonian.

Based upon growth studies, the following synonymies are suggested: E. constrictus Hall, E. ellipticus Miller, E. ovalis Hall, and E. subglobosus Miller are synonyms of E. crassus; E. clrodi Miller is a synonym of E. tuberculatus Miller & Dyer.

In modern vertebrates, particularly ectotherms in temperate climates, alternating periods of rapid and curtailed growth are often recorded by annual markings in bones and/or scales. By indicating the age at death of individual animals and the time of year when death occurred, such markings permit quantitative studies of growth rates and population dynamics of fossil vertebrates to be made, and aid in interpreting the origin of fossil deposits. Scales of fish collected from the varved, lacustrine Green River Formation (middle Eocene), southwestern Wyoming, show interruptions interpreted as annuli. The position of the most recent annulus relative to the edge of the scales indicates the season of death of individual fish. Perfectly preserved specimens died in early summer, whereas those that died at a later season are disarticulated and scattered. A likely explanation is that early summer mortalities coincided with the yearly summer maximum of carbonate precipitation in the lake, so that carcasses reaching the bottom at this time were protected from disruptive currents. Among a collection of early Pliocene vertebrates from a quarry in Nebraska are many specimens from which individual age and season of death can be determined. Annuli on gar scales and catfish vertebrae indicate that these fish were killed in winter. Among the mammals, merycodonts (aged by their horns and dentitions) and horses (aged by dentitions) apparently died in winter also. Unlike the Green River fish, which accumulated over thousands of years, the Nebraska occurrence probably resulted from a catastrophe.