The tetrad and companion cells, in association with the oesophagus and dorsal tract of tissue, of Tetradonema plicans, have been investigated in the host Bradysia paupera. The companion cells contain glycogen and mucopolysaccharide. The tetrad cells are rich in RNA and probably synthesize protein and secrete some as enzymes. The gut is functional and the oesophagus is in connexion with the tetrad. The dorsal tract is not connected with the oesophagus or tetrad and contains lipid with characteristics similar to that found in the eggs.

It is suggested that the tetrad and companion cells are homologous with the stichocytes present in the Trichuroidea and some members of the Mermithoidea.

I am grateful to Professor T. R. E. Southwood in whose department this work was carried out. I should like to thank Dr N. A. Croll for supervising and encouraging this work and Dr T. Jenkins for his helpful criticism and advice. The work was supported by the Agricultural Research Council.

An apparatus for the artificial feeding of tsetse has been described. Fresh defibrinated blood at 36 °C is circulated with a peristaltic pump through feeding capsules. The efficiency of the method has been compared with that of the Cockings' method and with the live host feeding. The flies exposed to feeding on this apparatus showed significantly higher feeding frequency than those allowed to feed on the Cockings' apparatus. Compared with the live host feeding, the feeding frequency for circulating blood method was slightly lower but the difference was not significant. Once the fly is induced to feed, it would in most cases complete engorgement unless disturbed. The rate of crop emptying in flies fed by the Cockings' method was abnormally slower as compared to those fed on the circulating blood. The present study indicates that for the successful artificial feeding of tsetse, it is essential to circulate the blood and regulate the temperature of the medium and the feeding surface to approximately 36 °C.

I wish to thank Mr S. B. Kutuza for technical help, and the Director of E.A.T.R.O. for permission to publish this work.

Adult specimens of Trochopus pini attach themselves by folding the disk-shaped haptor around the edge of the host's primary gill lamella. They are able to move from one place to another with the aid of disk-shaped pads on the head region.Ducts from the posterior median gland cells, which are extensive in the head region of the adult, open on the periphery of each disk and exude an adhesive secretion.

The oncomiracidium of T. pini is described and evidence is offered in support of direct invasion of the gills by these free-swimming larvae. Special attention is paid to the development of the head region of the parasite. Newly attached larvae lose their eye lenses and possibly their anterior median gland cells; their lateral head glands are well developed and most probably serve for attachment of the head region, since the posterior median gland cells are poorly developed and appear to have no ducts. Later the posterior median gland cells and their ducts increase in numbers and, in association with the pads on the head region, take over from the lateral glands the job of attaching the head region.

The paths of the tendons associated with the accessory sclerites on the haptor of T. pini indicate that this gill parasite has evolved relatively recently from the ancestors of benedeniid skin parasites.

I am most grateful to Commander J. Pareira, the Director of the Aquario Vasco da Gama in Lisbon where most of this work was done, and to the Staff of the Aquario and especially to Mr A. Ferreira. I am indebted to the Calouste Gulbenkian Foundation for a generous grant which made this visit possible. I am grateful also to the Directors and Staff of the Laboratoire Arago, Banyuls, France, the Station Biologique of the University of Bordeaux at Arcachon in France, and the Laboratory of the Marine Biological Association of the United Kingdom at Plymouth, England. I would like to thank the Royal Society for a grant from the Browne Fund which enabled me to work at Banyuls in 1962 and for a Pergamon Travel Award which made it possible to visit Arcachon in 1967.

Methotrexate inhibited the growth of D. viviparus by acting either on the moult from the third- to the fourth-stage larva or on the growth of fourth-stage larva itself. Despite these effects of the drug, the early developmental stages of the worms primed the immune mechanism of the host; the inductive phase of antibody production was completed in less than 5 days. An effective immune response, however, was blocked by the continued presence of methotrexate which resulted in delayed elimination of a superinfection and the absence of anaphylactic antibody in the serum. Immune elimination appeared to require a quantitative threshold of worm antigen and host antibody.

This work was made possible by the generous supply of infective helminth material from Messrs Allen and Hanbury Ltd, Ware, Herts. I should also like to thank Dr J. H. Humphrey and Dr S. R. Smithers of N.I.M.R. for their helpful discussions.

After 24 h desiccation at 28–29 °C and approximately 50 % relative humidity about 70 % of B. microplus larvae drank sufficient solution containing 1·5 mCi of 32P per ml for their movements on the host to be followed with a portable G-M counter. Reduced viability of the labelled ticks could not be detected when they were kept for 24 h on the host and either allowed to feed or prevented from doing so. When feeding was continued for 72 h or when larvae were incubated for 3–4 weeks, adverse effects were found, the severity of which could not be directly related to the amount of 32P label. Desiccation, handling of larvae and exposure to radiation in the drinking solution may have contributed to their reduced viability.

Since resistance to the tick is manifest within the first 24 h after infestation labelled larvae could be used to compare development on different hosts during this time. Even after feeding for 72 h, larvae were more adversely affected by host resistance than by labelling damage.

We would like to thank Mr C. A. Schuntner for helpful advice and Mrs J.Galligan for technical assistance. The financial support of the Australian Meat Research Committee is gratefully acknowledged. Messrs D. Hancox and S. R.Henderson assisted with some of the experiments.

Haliotrema balisticus attaches itself between the secondary gill lamellae of its host, Balistes caprisctis (= carolinensis), by means of a pair of ventral hamuli, which point ventrolaterally, and a pair of dorsal hamuli which point in a dorsal direction. The ventral hamuli, each of which pivots at the end of a ventral connecting bar, are operated by a pair of extrinsic muscles. The tendinous regions of these muscles pass through fibrous loops attached to the ventral hamuli and then join each other in the mid-line. Such an arrangement has a high mechanical advantage which may reduce the expenditure of energy required to maintain the hamuli in position in the gill tissue in the face of a strong gillventilating current and may also preserve the mobility of the free region of the body.

The dorsal hamuli are not operated by the muscles and tendons associated with the ventral hamuli.

Marginal hooklets and glands play subsidiary roles in attachment.

I wish to express my thanks to Commander J. Parreira, Director of the Aquario Vasco da Gama in Lisbon, and to the Staff of the Aquario, in particular to Mr A. Ferreira, for their kind assistance and hospitality. I am also grateful to Dr Maria Monteiro and to Dr Maria Baptista. I am indebted to the Calouste Gulbenkian Foundation for a grant which made my visit to Portugal possible.

The genitalia of the Pupipara are markedly modified from those of other higher Cyclorrhapha and the nomenclature used is mainly descriptive and controversial; different terminologies have been used by the various authors.

A drug-resistant strain of T. vivax was transmitted through a series of four duikers (S. grimmia) and a gazelle (G. rufifrons) by G. palpalis and G. tachinoides. The cyclical transmission of the strain in cattle, started in earlier work, was continued for control purposes. Infections with the drug-resistant strain in the antelope were of the same type as those produced by field strains of T. vivax in the same host species, and infection rates in tsetse flies fed at early stages of infection on the antelope were similar to those found in flies fed on cattle infected with the drug-resistant strain. The strain retained resistance to diminazene aceturate at a dose rate of 7-0 mg/kg and quinapyramine sulphate at 5'0 mg/kg in the absence of the drugs for 7 months in tsetse flies and antelope, and for a total period of 29 months in tsetse flies and cattle.

We wish to thank Dr A. A. Amodu, Director of the Nigerian Institute for Trypanosomiasis Research, for permission to publish this paper.