A review of the postembryonic development of millipedes (Diplopoda) is given, based mainly on published information. Original observations are, however, also included. Millipedes hatching from the pupoid usually have three pairs of legs; during their postembryonic growth they acquire more segments and more legs. This process is known as anamorphosis. Three types of anamorphosis are recognized. In euanamorphosis, every moult is accompanied by addition of new segments, even after the attainment of sexual maturity. In hemianamorphosis, the addition of new segments goes on until a certain stadium, and further moults take place without addition of segments. In teloanamorphosis, the addition of segments stops at a certain stadium (the adult, and ultimate, stadium) after which no further moults occur. Available information on anamorphosis in each of the millipede orders is reviewed. General patterns are emphasized, but variations are also considered in detail. It is shown that the so‐called ‘law of anamorphosis’ is valid only for the ‘ring‐forming’ millipedes (Merocheta and Juliformia) in which tergites, pleurites, and sternites of each diplosegment are firmly fused into a complete ‘ring’, and for some other forms (Polyzoniida, Chordeumatida), where there is a constant relationship between rings and legs. The chapter on the order Julida is particularly detailed and includes discussions of patterns in the variation and a section on periodomorphosis. The general chapter on developmental patterns includes inter alia an interpretation of the variations in millipede anamorphosis in terms of the ‘biometabolic modi’ of Remane. The hypothetical ancestral millipede is shown to have developed by hemianamorphosis. Euanamorphosis was acquired by the ancestral species of Helminthomorpha. Within this clade, Chordeumatida and Merocheta have secondarily become teloanamorphic, whereas some Juliformia seem to have returned to hemianamorphosis. The contrasting principles of elongation and contraction, subject of much debate among diplopodologists, are shown both to have played a role in the course of millipede evolution.

The density and surface activity of the millipedes in a quarter of an acre of a sycamore ash wood are described. Seven species were extracted by Tullgren funnels from samples of soil and litter over five years and were also caught in pitfall traps during a further two years; four other species occurred occasionally in the traps. Each square metre of the site supported 100 individuals over the winter, rising to 300 in the summer. Of these, 85 % belonged to three species, Iulus scandinavius, Polydesmus angustus and P. denticulatus. Male I. scandinavius become adult in either the ninth, tenth or eleventh stadium; females in the tenth and eleventh. Eggs are laid in spring and these take three years to become adults which breed and then die. The majority of Polydesmus spp. in the samples are young belonging to the first six of the eight free‐living stadia. The adults fell into the traps in the summer and newly emerged young appear in the samples at this time. They overwinter in their first year mainly as fifth stadia; some might reach maturity (eighth stadia) in the summer following, but it is not certain that they could breed at this time. The pattern of dispersion of lulus is fairly even and is correlated with the distribution of leaf litter but the Polydesmus spp. are highly aggregated. All stadia of lulus fall into the traps but only the last two of Polydesmus. The aggregation of Polydesmus spp. appears to be correlated with the relative inactivity of the younger stadia. The estimates of density of Polydesmus spp. are unreliable because of their aggregation but those of lulus have determinable limits and it is possible to derive rough though meaningful standing crop, production and life‐table data. The overwintering standing crop of lulus consists of the survivors of three generations of 5, 2.2 and 1.4% of the original eggs laid; it has a fresh weight of about 1.25 g and a production in the order of 1.5–2.5 g/m2.

A population of C. nitidus in a deciduous wood at Milldale, Derbyshire included adult males with 36 to 54 podous segments but the highest number of segments possessed by immature males was 38. The occcurrence of males in successive stadia is contrasted with the alternation of copulatory and intercalary males in some other juliform millipedes. Samples were taken to determine the course of post‐embryonic development and life‐history in the field. Stadial determination by counting rows of ocelli was not possible but a separation was made by analysis of dimensions. These dimensions and segmental formulae are compared with those of German examples. No intercalary or other modified post‐imaginal male stages were present. From a comparison of the distribution of stadia in five samples it appears that males mature in three and females in four years. Males survive a further four to five, females a further three years after first achieving maturity. The significance of the special and general life‐history characters of the species are discussed.

In some iulids females can moult several times after first attaining maturity but males cannot. Males of Tachypodoiulus niger (Leach) are exceptional in that they can moult, but they lose their functional intromittent organs in the process, regaining them after a second moult. The process can be repeated, a series of functional males alternating with nonfunctional males. This extension of the life and the consequent increase in number of segments led Verhoeff (1928) to postulate that the ancestral millipede was short‐bodied and that many‐segmented forms were derived from it. We think that this extension of life is of ecological rather than phylogenetic significance–adapting the species to disperse widely to scattered habitats. In particular, the extension of life of males as well as of females may ensure a reasonable sex ratio in those areas where the species is least dense. In Britain, Tachypodoiulus niger lays eggs in spring which reach the fourth and fifth stadia by their first winter and the seventh, eighth and ninth stadia by their second winter. Maturity is usually attained by males in the eighth stadium but occasional specimens mature at stadium seven and others defer maturity until the ninth. Both sexes can proceed to the fourteenth stadium but adults in Britain usually belong to two generations of two and three years old, divided mainly between the eighth, ninth and tenth stadia. Details are given of a large collection made by the late Dr Scott of animals taking refuge in his house at Henley‐on‐Thames. Animals in Britain appear to have a similar life history to those in Germany as described by Verhoeff, but effective comparison has depended on correcting and re‐interpreting some of Verhoeff's data.

No abstract