1. The histology of the venom apparatus and the histochemistry of the venom glands of 14 species of fishes belonging to six families have been described. Data additional to those already existing on the venom apparatus of another three species, belonging to three families, have been collected. Histological examination of a further three species showed that they lack a venom apparatus. These are Paraploactis trachyderma (Aploactidae), Halophryne diemensis (Batrachoididae) and Acanthurus xanthopterus (Acanthuridae).
2. The scatophagids, Scatophagus argus and Selenotoca multifasciata possess venom glands which are wedge-shaped epidermal thickenings occupying paired spine grooves and contain aggregations of unicellular glands producing proteinaceous secretions. The venom cells possess unique rod-like crystals of enigmatic nature. The lengths of the venom glands, both absolutely and relative to spine length, decrease with increasing size of fish.
3. All fish venom glands are aggregations of unicellular glands, though they function as multicellular entities. They produce venom by holocrine secretion.
4. Supporting cells, similar to epidermal undifferentiated cells, are a constant feature of all fish venom glands. The fundamental reasons for their presence are discussed in relation to the derivation of fish venom glands from the epidermis.
5. A scorpaenid, Notesthes robusta, a synancejid, Inimicus barbatus and five Teuthis species have been shown to elaborate venom granules similar to those of the stonefish Synanceja trachynis.
6. The mode of regeneration of scorpaenid venom glands has been demonstrated. Regeneration is accomplished by inward growth of a bridge of epidermal cells from the epidermis of the spine integumentary sheath. Such a bridge is a permanent feature at the distal ends of venom glands in Pterois volitans. Similar bridges occur along the entire venom gland length in some catfishes (Plotosus lineatus, Euristhmus lepturus and Cnidoglanis macrocephalus).
7. All teleost venoms are basic proteinaceous secretions containing few lipids or mucosubstances.
8. Certain factors common to all venomous fishes suggest the following procedures for histological studies: to prevent autolytic disruption of venom cells, freshly killed specimens should be examined; the proteinaceous, often granular secretions of fish venom glands are suitably fixed by Baker's formol calcium and acidic formalin mixtures should be avoided; nitric acid mixtures are unsuitable decalcifying agents and in the present study, Gooding and Stewart's decalcifying mixture was found to be satisfactory.
9. Each venom apparatus studied was a defence mechanism. If a venomous fish is molested, its venomous spines are erected and envenomation occurs when the spines penetrate the flesh of an animal. Penetration usually occurs when an animal impales itself on the spines but movements of the venomous fish can facilitate penetration as in the case of most catfishes and Pterois.
10. Danger to man from any venomous fish can be correlated with a complex of factors including absolute size of venom glands, production of venom granules (except in catfishes), traumatising ability of spines (serrate in catfishes, needle-like in other species) and habits of the species. In catfish species, the greater the aggregation of venom cells in the vicinity of the traumatising spine, the greater the danger to man.
11. The epidermis of five catfishes, an aploactid, a scorpaenid, a batrachoidid and an acanthurid has been described. Each species except the acanthurid, Acanthurus xanthopterus, possesses glandular epidermal cells in addition to mucus-secreting cells. Their non-mucoid secretions may be distasteful to predators or toxic to organisms capable of invading the integument. Possible evolutionary development of venom glands from such epidermal glands is discussed.
12. Methanol extracts of skin of the catfish, Tandanus tandanus and one aqueous extract of Cnidoglanis macrocephalus skin contained compounds which stimulated barnacle muscle preparations, causing increased amplitude of contractions at heightened tone.
13. Anatomical, histological, histochemical and ultrastructural studies have been made of the axillary glands of a catfish, Cnidoglanis macrocephalus. The clavate cells of which the glands are principally composed differ somewhat from clavate cells of the general body epidermis of the species. Also, these axillary glands are remarkable in that their diameter is approximately five times that of catfish axillary glands previously described.
14. Axillary gland secretion of C. macrocephalus elicited activity in rat uterus, rat ileum and barnacle muscle preparations. It did not affect the rat phrenic nerve-diaphragm preparation.
15. Compounds possessing biological activity have been isolated by paper chromatography from NaCl and acetone extracts of C. macrocephalus axillary gland secretion. They appear to be phenolic tryptamines.