Although our knowledge of octocorals has increased greatly in the last half a century, notably aided by the advances in laboratory and underwater technologies, there is still much we need to know, and much we will probably never know due to the absence of a good fossil record of these animals. The latter is all obvious factor contributing to the deficiency of the knowledge needed for a satisfactory division of the Order Alcyonacea into suborders; there being consensus only about the holaxonian (see page 58) arid the calcaxonian (page 62) gorgonians. Time has seen the other families subjected to various groupings, at various levels of classification, but newly discovered material, and the re-evaluation of old findings, has shown such overlap between these groups that their maintenance as suborders is difficult to justify. We include them here as "groups" only. The names are often useful when talking in generalities, but using them as discrete categories should be viewed with some caution.

A few authors still group the following families of soft corals under the name Stolonifera. The name was originally proposed as an order to cover all those genera having species where the polyps arise separately from ribbon-like stolons, and are not united side by side in a common fleshy mass. However, a review of all the genera that grow in a manner similar to this shows there is a clear series of transitional forms between those where the polyps are not united, and those where they are united in a common coenenchyme. Stolons can merge to form membranes, membranes can grow quite thick, and `separate` polyps can show various amounts of fusion. The decision whether to categorise a particular genus as a stoloniferan becomes so subjective that the name plainly has limited classificatory value, except perhaps to indicate the growth form of some of the more primitive genera.

Colonies in this family are usually quite small, in some cases minute. They consist of cylindrical or bluntly conical polyps usually joined only at their bases by reticulating stolons, which may coalesce into thin membranous expansions. In some genera, tall cylindrical polyps develop long secondary polyps that resemble branches. in a few instances, polyps are also connected by extra, transverse, bar-like stolons above the basement layer. There is little overlap between the shallow-water Indo-West-Pacific genera in this family and those in other families. The only similar genus is Comularia, but it grows in European waters, and has a thicker chitinous envelope around each polyp. This thick cuticle forms a theca-like cup into which the polyp can withdraw, and so Comularia is not included in this family.

A thin, soft, brownish, transparent covering (called the perisarc), enveloping the stolons and the lower parts of the polyps, is characteristic for most genera in this family (it is worth noting that probably all soft corals have a perisarc coating near their very base, or at least between their base and the substrate upon which they are growing). However, the thin perisarc can easily be overlooked, and in many cases it is extremely difficult to detect without histological preparations. Sometimes, light reflecting from wrinkles, or the presence of detritus and foreign organisms adhering to the perisarc surface, is the only clue to its presence. Because there are many other morphological characters that can be used to distinguish the genera we have illustrated, searching for the perisarc is not necessary.

Nearly all of the species in this family have sclerites, and it is not unusual for these to be fused into clumps or tubes. Sclerites include such forms as smooth branched rods, and prickly or tuberculate 6-radiates, spindles and platelets.

Tubipora is the only genus in this family. Colonies are hemispherical and massive, to thick and encrusting, and consist of vertical, red, hard calcareous tubes connected at various levels by horizontal, stolonic platforms. Colonies can be up to 50 cm in diameter. Each tube is formed and occupied by a single polyp, which is connected to the other polyps in the colony by canals inside the horizontal plates. Stolons and non-extendable parts of the polyps are covered in a thin, soft perisarc.

Although the name Alcyoniina appears now and then in textbooks and sites on the world wide web, it is rarely seen in modern literature dealing with the identification of octocorals. The category is meant to include all those soft corals in which groups or all of the polyps are united side by side in a common, fleshy mass, without a supporting axis of horny and/or calcareous material. However, it is now obvious that a complete series of intermediate forms link the primitive soft corals in the Stolonifera group to a number of the genera included in the Alcyoniina and Scleraxonia group (page 53).

Colonies in this family are the dominant reef dwelling octocorals in the Indo-West- Pacific. Their growth form is often termed "massive", referring to the fact that their dimensions are similar in all directions, and that the polyps are united to form fleshy masses. This is, of course, a generalisation, and there are species that grow in the shape of cigars or carrots, others form extensive encrusting mats several meters across and only a couple of centimetres thick, and one genus grows as membranous expansions only millimetres thick.

In most colonies, there is a bare basal section, the stalk or trunk, and an upper, polyp-baring part, divided into lobes, ridges or short branches. In some, the upper surface is flat, undulate, or loosely pleated around the margin. Massive or carrot-like colonies cut in half will reveal many very long canals inside, which are the gastric cavities of numerous primary autozooid polyps extending from the colony base to the uppermost regions. On the colony surface, siphonozooid polyps may be present amongst the autozooids.

Sclerites include tuberculate or prickly spindles, clubs, 6- or 8-radiates, ovals and dumbbells. The interior of a colony may be compressible and jelly-like if the sclerite content is low, and ridged and solid if it is high.

An often bewildering set of genera are grouped in this family, some probably erroneously. Most can be described as bushy, globe-shaped or arborescent in growth form, while a few others are massive, and one genus consists of numerous finger like lobes united by a common base. Many genera contain highly coloured species.

In most cases the polyps, singly or in small clusters, are more or less restricted to the upper and outer twigs or branches. In a few cases, polyps grow directly on main branches. If an arborescent colony is cut longitudinally, a small number of broad primary polyp canals can be seen, which split into groups that extend up into the distal lobes and branches. The canal walls are generally thin, with few sclerites, permitting colonies to easily inflate with water and dramatically increase their size.

Although some genera are soft and floppy, nephtheids are commonly known for their rough or often distinctly prickly feel. In the upper colony region, this is caused by the strong, protective sclerite armature of the polyps, which may project for several millimetres beyond the polyp head. In the stem and branches, the sandpaper-like texture can be attributed to numerous, strongly sculptured, spiny sclerites in the surface layer. Sclerite forms include prickly needles, leafy clubs, irregular shaped spiky forms, and tuberculate and thorny spindles, often extensively ornamented along one side.

Three genera (Nephthea, Litophyton and Stereonephthya) have been the source of much confusion for octocoral taxonomists. They contain a complex array of sclerite forms, of sclerite arrangements in the polyps, and of polyp distribution. Fortunately, a complete revision of this taxonomic chaos by our friend and colleague Leen van Ofwegen, of the Nationaal Natuurhistorisch Museum in the Netherlands, is nearing completion after years of research. The description of these genera has been kindly given to us by Leen, with the proviso that it is possible the details may be altered when the research is finalised, as there are still a number of problems to be solved.