Glycerids are carnivorous polychaetes and Glycera alba, for example, has been shown to feed mainly on errant polychaetes and small crustaceans, chiefly amphipods (Ockelmann & Vahl, 1970). Glycerids inject their prey with a venom containing a toxin and proteinases (Michel & Keil, 1975), produced in discrete glands (Michel, 1966, 1970), by means of four jaws placed symmetrically around the tip of the muscular proboscis. The distal, free half of each jaw is slender and curved, the basal section much broader and enclosing the ‘pulp cavity’. Attached to one side there is a jaw support (‘aileron’), usually V-shaped with unequal prongs but the shape varies considerably between species (see Wolf, 1977). The jaws are operated by strong longitudinal muscles attached to the jaw bases and supports (see Gravier, 1898). Nearly the whole length of the jaw is traversed by a fine venom canal, its position marked on the surface by a linear series of small pores (Fig. 1 A) through which the venom is probably emitted (Charletta & Boyer, 1974). In colour the jaws are a uniform shiny jet-black.

Facilities for keeping marine fish for experimental purposes or for display to the public are usually run on either an open system with direct supplies of water from the sea, as in the Bergen Aquarium (Rollefsen, 1962), or on a mainly closed system in which the sea water is continually recirculated (Saeki, 1958). The pros and cons of each system are well known and have been discussed in the published papers of the 1960 1st International Congress of Aquarology held in Monaco (Comité Scientifique et Technique du Congrés International d'Aquariologie, 1962–63) and in the report of the ad hoc meeting in 1975 on design and practical operation of research aquarium systems held at Texel, The Netherlands (International Council for the Exploration of the Sea, 1975). A combination of the two types of system is often used in research institutions and series of papers on sea-water aquarium designs and methods of operation have been collected by Clark & Clark (1964).

The possibility of dissolved organic compounds acting as complexing or chelating agents in natural waters has received considerable attention in the last two to three decades. Stumm & Morgan (1970) have expressed doubts about the existence of humicmetal ion complexes in natural waters. Strickland (1972) has also stated that although the addition of chelating agents to sea water often improved the growth of phytoplankton, there was little evidence that the function of dissolved organic matter in oceans and lakes was to complex metals so as to increase or decrease their availability to phytoplankton. Strickland argued that even if all dissolved organic carbon were present as a compound of strong complexing ability it would not be able to compete for most metals with inorganic ligands such as chloride, sulphate and hydroxide. However, there is an increasing amount of data indicating that metals in natural waters may exist in chelated forms with dissolved organic matter (see, for example, Davey, Morgan & Erickson, 1973; Chau & Lum-Shue-Chan, 1974).

Heavy metals, as well as affecting the growth rate and photosynthesis of algae, have a considerable effect on the size and morphology of algal cells (see, for example, Erickson, 1972; Nuzzi, 1972; Davies, 1974; Kayser, 1976). Changes in the colour of algae in the presence of heavy metals have also been observed (Greenfield, 1942). At metal concentrations resulting in total growth inhibition of algae the cells, although abnormal, are not necessarily dead (see, for example, Nielsen & Kamp-Nielsen, 1970; Erickson, 1972; Nuzzi, 1972; Blankenship & Wilbur, 1975).

In heavy-metal toxicity studies with Ditylum brightwellii (West) Grunow (Canterford & Canterford, 1979) it was observed that low concentrations of ethylenediaminetetraacetic acid (EDTA) and toxic concentrations of several heavy metals resulted in the formation of abnormal cells. In the present paper factors affecting the formation of abnormal cells are discussed. Experiments were carried out to establish whether or not these abnormal cells could be regenerated.

In making benthic surveys it is often difficult to extract all the animals from the various sediments in good condition. Even the IBP Handbook (Holme & Mclntyre, 1971) does not give comprehensive details of methods for extracting animals from the complete range of sediment types. The greatest difficulty is presented by clays and compacted silt. In recent work we have had very satisfactory results by using the following procedure. Core and grab samples taken in the field are stored in seal-tight buckets for processing in the laboratory. In the case of grab samples taken from boats, preliminary sieving is performed on site and only the resistant sediments returned to the laboratory. The samples are broken into pieces of approximately 100 cm3 or less and stored in 5 % formaldehyde solution using a volume about three times that of the sediment. This serves to fix and preserve the fauna and also helps in the breakdown of clays. At this stage samples may be stored for a long time. Wherever layers of soft mud are found among the resistant material they are washed off, sieved and the fauna preserved separately. The resistant samples are then frozen in a domestic freezer and left for 24 h before thawing. The freezing is very effective in breaking down resistant sediments into small particles, thus releasing the animals, which can readily be removed by elutriation. Normally, most animals seem to be adversely affected by freezing in much the same way as some vegetables lose their consistency when frozen (an exception to this are members of the polychaete family Phyllodocidae). However, the prior treatment with formalin seems to protect the tissues from damage by freezing.