Modular organization of colonial hydroids is based on cyclic morphogenesis during growth of their body. In many thecate hydroids (Hydrozoa: Leptomedusae) the shoots of the colony consist of a few distinct elements and possess complex spatial organization. In most cases, the evolutionary sequence of morphogenetic modifications that led to present-day organization of shoots is obscure and not obvious. One of the approaches that allow getting insight into the morphogenetic evolution in colonial thecate hydroids is to analyse the spectrum of different minor morphotypes presented in the population of the certain species. In our opinion, some rare morphotypes allow understanding and reconstructing the scenario of morphogenetic evolution of species under consideration. We describe the application of such an approach for reconstruction of the morphogenetic evolution of Dynamena pumila (L.) (Sertulariidae) with some additional conclusions.

Cytomorphological characters of Polypodium hydriforme—a unique cnidarian, adapted to intracellular parasitism inside oocytes of Acipenseriform fish—are reviewed and analysed. Polypodium's unique characters, those shared with the Myxozoa, and the characters shared with bilateral triploblastic animals are listed. The set of unique features has allowed us to allocate Polypodium to a separate cnidarian class Polypodiozoa Raikova, 1988. The combination of characters displayed by P. hydriforme suggests that this animal is not an aberrant cnidarian, as it was formerly believed, but a relic of a major cnidarian group (class Polypodiozoa) that has undergone reduction due to adaptation to parasitism.

The simplified text-book view holds that hydroids' soft body is composed of a branched double-layered tube, whose wall consists of two epithelial layers (the inner gastrodermis and the outer epidermis) separated by the mesoglea. Some hydroids are characterized by large, complex colonies and likely an even more complicated inner organization. By using three species from the thecate hydroids of the family Sertulariidae we investigated the soft body structure of such hydroids. The anatomical study revealed some new features of colonial hydroids. The double layered coenosarc fills the perisarc (outer skeleton) tube only at the endings of the branched colony. More proximally, the coenosarc tube becomes narrower and a thin epidermal lining covers the inner surface of the perisarc tube. In some species the soft tissues of the shoots form a network of anastomosing canals. The canals are formed by the gastrodermal epithelium and they are embedded in epidermal tissue. In the upper part of the shoot, these canals are located at the periphery, along the inner surface of the perisarc. In more proximal regions of the stem, the whole lumen of the perisarc tube can be occupied by gastrodermal canals; the canals are enclosed in a parenchyma-like epidermal tissue. The organization of the soft tissue in these thecate hydroids is a striking example of structural complexity that does not contravene the limits of the ground plan of the phylum.

Hydroids can establish symbiotic relationships with most marine phyla. Almost entire genera or even families are associated with specific groups (e.g. Hydractiniidae and Cytaeididae with gastropods and hermit crabs, Zancleidae with bryozoans, Dipurena with sponges, Ralpharia with octocorals, Eugymnanthea with bivalves, Proboscidactyla and Teissiera with serpulids, Bythotiara with tunicates). Generally, the symbiotic groups belong to the Anthomedusae that, due to the absence of theca, are more plastic in establishing trophic relationships with the hosts. Nevertheless a number of scattered species, mainly Leptomedusae, are strictly associated to algae or sea grasses: in these cases no evident morphological or behavioural adaptations were observed. In animal symbiosis several unrelated symbiotic species show polymorphic colonies or a strong reduction in number and/or size of the tentacles, which are sometimes completely lost. Moreover, these symbiotic species may lack perisarc even in the hydrorhiza.

In this paper we summarize the morphological and behavioural adaptations of symbiotic species suggesting that the described aptitude of hydroids to establish relationships with other organisms is not only the result but also the source of the evolutionary radiation of this group.

Study of living colonies of Salacia tetracythara Lamouroux, 1816, led to an accurate description of the hydranths and adjoining tissues. A mantle is reported for the first time for this species, lining stem, hydrocladia and hydrotheca. An adcauline nematocyst clump located in the median part of the column, composed of large microbasic mastigophores, is also reported as a new structure for sertulariids. The hypothetical modified tentacles noted by Billard (1925) are confirmed and re-described. Among the 22–28 tentacles of the hydranth in a circlet, two exhibit nematocyst-bearing organs connected at base. These organs are massive, racket-shaped, refringent, and composed of large microbasic mastigophores. They do not affect the tentacles in respect of their general shape and behaviour. One nematocyst-bearing organ can sometimes be linked to more than one tentacle depending on its size. The abcauline position of the two modified tentacles imparts a strong bilateral symmetry to the hydranth. This location explains why only one modified tentacle was previously described from preserved material, the second being hidden by the first when hydranths are retracted inside the hydrotheca. The modified tentacles of S. tetracythara are compared with the nematodactyls of the genus Nemalecium, with the ligula found in several sertulariid species, and with nematophores described in other thecate families. Beside remarks on species taxonomy and distribution, incitation is given to the study, with emphasis especially on hydranths.

The study of living Halecium halecinum colonies revealed a new case of zooid polymorphism. Besides the ordinary hydranth, the polyp devoted to feeding, this species is provided with a second kind of polyp, differing only slightly in structure and morphology but most conspicuously in its behaviour. It is named ‘extensible polyp’ in reference to its great extensibility and the resulting filiform shape. There are slight differences in the tentacles: lower number, shorter length, thicker diameter, and the tip slightly swollen and rounded instead of tapering. Their large microbasic mastigophores are abundant and evenly distributed, while the hydranth has a few large ones only on the oral side but has otherwise numerous small ones. When extended and at rest, the tubular column is much longer than that of the hydranth and not delimited from the head of the polyp by a bulge followed by a constriction. Behavioural differences are its capacity to coil and bend during extension and thus being able to move in all directions and exploring a large volume of seawater, and also its ability to produce regional swellings (peristalsis) and to contract by folding and bulging though still extended. Besides a probable role in defence, the extensible polyp exhibits an excretory function and it could also have sensory functions. The extensible polyp type is not classified as a nematophore because it has a functional gastrovascular cavity and a mouth. Polyp dimorphism (hydranth/extensible polyp) is reported in one more halecid and two sertularids.

A creeping movement, which often results in clockwise or counterclockwise bodily rotation, is described for ordinary zooids of two species of bivalve-inhabiting hydrozoans, Eutima japonica Uchida and Eugymnanthea japonica Kubota, attached to artificial hard substrata (polystyrene and glass) in the laboratory. Time-lapse video recordings furthermore showed that the non-transparent half of the flat, ellipsoidal pedal disc always faces forward during this motion, with no lateral or backward creeping during either day or night. Because an ordinary zooid generally crept in a circle of small radius, the zooid's body rotated as the creeping proceeded. One full rotation of a zooid of either species took about 5–6 hours on average at 21–25°C. The direction of rotation of an individual zooid was nearly constant irrespective of developmental state, this being true for zooids with or without a medusa bud on the hydrocaulus or a daughter zooid on the pedal disc, those with a normal or a double hydranth, or those consisting of a bare hydrocaulus. Asexually produced tiny daughter zooids quickly became well separated from their mother zooids. The attachment site of the latter did not shift much as they rotated, but daughter zooids and other tiny free zooids creep along a nearly straight line, thereby markedly shifting their position. This may contribute to the prevention of crowding of zooids in the restricted space between soft body parts in the host bivalve's mantle cavity.

The auto-fluorescence patterns of the medusae observed under a fluorescent microscope with blue light excitation allows to distinguish two species of Eugymnanthea, this even when they are still attached to the hydroid as small medusa buds despite the occurrence of a sex-dependant pattern in E. japonica. A total of four distribution patterns of green fluorescence, including non-fluorescence, could be found. Three of them are found in E. japonica, called ‘subumbrellar fluorescence type’ except for non-fluorescence, while another type is found in E. inquilina, called ‘umbrellar margin fluorescence type’. During the short life of the medusa the latter type remained invariable for up to six days in E. inquilina, while the pattern observed for up to seven days in E. japonica changed sometimes, but it always remained distinguishable from the pattern found in E. inquilina. Therefore, the fluorescence pattern is a reliable taxonomic character. Fluorescence was not found in unfertilized eggs, planulae 2–8 days old, parthenogenetically produced larvae, or in the hydroids of the two species. The auto-fluorescent and possible bioluminescent tissues of these Eugymnanthea medusae could have some unknown biological significance.